From ab3739219e452b73496021d3503358e74f1ff0b7 Mon Sep 17 00:00:00 2001
From: jialin <jialinkuang@126.com>
Date: Thu, 18 Jun 2026 15:42:58 +0800
Subject: [PATCH] chore: update vllm, sglang, preset argument options

---
 src/pages/llmodels/config/index.ts            |    2 +-
 .../constants/backend-parameters/sglang.ts    |  727 ++++++--
 .../constants/backend-parameters/vllm.ts      | 1502 ++++++++++++-----
 3 files changed, 1663 insertions(+), 568 deletions(-)

diff --git a/src/pages/llmodels/config/index.ts b/src/pages/llmodels/config/index.ts
index 478015e3..17b3a4d6 100644
--- a/src/pages/llmodels/config/index.ts
+++ b/src/pages/llmodels/config/index.ts
@@ -455,7 +455,7 @@ export const getBackendParamsTips = (backend: string) => {
     return {
       backend: 'SGLang',
       releases: '',
-      link: 'https://docs.sglang.ai/advanced_features/server_arguments.html',
+      link: 'https://docs.sglang.io/docs/advanced_features/server_arguments',
       version: 'v0.5.4'
     };
   }
diff --git a/src/pages/llmodels/constants/backend-parameters/sglang.ts b/src/pages/llmodels/constants/backend-parameters/sglang.ts
index f12238ac..f2958f14 100644
--- a/src/pages/llmodels/constants/backend-parameters/sglang.ts
+++ b/src/pages/llmodels/constants/backend-parameters/sglang.ts
@@ -1,5 +1,12 @@
 import { BackendParameter } from './index';
 
+// Generated from the SGLang `launch_server` server arguments reference:
+// https://docs.sglang.io/docs/advanced_features/server_arguments
+// Grouped by the doc's sections. `options` lists the fixed choice values
+// when the flag accepts a restricted set (sourced from the doc's Options column).
+// Internal/experimental sections (Ktransformers, Diffusion LLM, Forward hooks,
+// msProbe dump) and deprecated arguments are omitted.
+
 const options: BackendParameter[] = [
   // Model and tokenizer
   {
@@ -45,8 +52,12 @@ const options: BackendParameter[] = [
       'gguf',
       'bitsandbytes',
       'layered',
+      'flash_rl',
       'remote',
-      'remote_instance'
+      'remote_instance',
+      'fastsafetensors',
+      'private',
+      'runai_streamer'
     ]
   },
   {
@@ -85,11 +96,26 @@ const options: BackendParameter[] = [
     options: []
   },
   // HTTP server
+  {
+    label: '--host',
+    value: '--host',
+    options: []
+  },
   {
     label: '--port',
     value: '--port',
     options: []
   },
+  {
+    label: '--fastapi-root-path',
+    value: '--fastapi-root-path',
+    options: []
+  },
+  {
+    label: '--grpc-mode',
+    value: '--grpc-mode',
+    options: []
+  },
   {
     label: '--skip-server-warmup',
     value: '--skip-server-warmup',
@@ -105,6 +131,11 @@ const options: BackendParameter[] = [
     value: '--nccl-port',
     options: []
   },
+  {
+    label: '--checkpoint-engine-wait-weights-before-ready',
+    value: '--checkpoint-engine-wait-weights-before-ready',
+    options: []
+  },
   // Quantization and data type
   {
     label: '--dtype',
@@ -124,6 +155,7 @@ const options: BackendParameter[] = [
       'bitsandbytes',
       'gguf',
       'modelopt',
+      'modelopt_fp8',
       'modelopt_fp4',
       'petit_nvfp4',
       'w8a8_int8',
@@ -131,7 +163,12 @@ const options: BackendParameter[] = [
       'moe_wna16',
       'qoq',
       'w4afp8',
-      'mxfp4'
+      'mxfp4',
+      'mxfp8',
+      'auto-round',
+      'compressed-tensors',
+      'modelslim',
+      'quark_int4fp8_moe'
     ]
   },
   {
@@ -139,6 +176,16 @@ const options: BackendParameter[] = [
     value: '--quantization-param-path',
     options: []
   },
+  {
+    label: '--kv-cache-dtype',
+    value: '--kv-cache-dtype',
+    options: ['auto', 'fp8_e5m2', 'fp8_e4m3', 'bf16', 'bfloat16', 'fp4_e2m1']
+  },
+  {
+    label: '--enable-fp32-lm-head',
+    value: '--enable-fp32-lm-head',
+    options: []
+  },
   {
     label: '--modelopt-quant',
     value: '--modelopt-quant',
@@ -155,13 +202,23 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--kv-cache-dtype',
-    value: '--kv-cache-dtype',
-    options: ['auto', 'fp8_e5m2', 'fp8_e4m3']
+    label: '--modelopt-export-path',
+    value: '--modelopt-export-path',
+    options: []
   },
   {
-    label: '--enable-fp32-lm-head',
-    value: '--enable-fp32-lm-head',
+    label: '--quantize-and-serve',
+    value: '--quantize-and-serve',
+    options: []
+  },
+  {
+    label: '--rl-quant-profile',
+    value: '--rl-quant-profile',
+    options: []
+  },
+  {
+    label: '--enable-quant-communications',
+    value: '--enable-quant-communications',
     options: []
   },
   // Memory and scheduling
@@ -190,6 +247,16 @@ const options: BackendParameter[] = [
     value: '--chunked-prefill-size',
     options: []
   },
+  {
+    label: '--prefill-max-requests',
+    value: '--prefill-max-requests',
+    options: []
+  },
+  {
+    label: '--enable-dynamic-chunking',
+    value: '--enable-dynamic-chunking',
+    options: []
+  },
   {
     label: '--max-prefill-tokens',
     value: '--max-prefill-tokens',
@@ -198,13 +265,26 @@ const options: BackendParameter[] = [
   {
     label: '--schedule-policy',
     value: '--schedule-policy',
-    options: ['lpm', 'random', 'fcfs', 'dfs-weight', 'lof', 'priority']
+    options: [
+      'lpm',
+      'random',
+      'fcfs',
+      'dfs-weight',
+      'lof',
+      'priority',
+      'routing-key'
+    ]
   },
   {
     label: '--enable-priority-scheduling',
     value: '--enable-priority-scheduling',
     options: []
   },
+  {
+    label: '--abort-on-priority-when-disabled',
+    value: '--abort-on-priority-when-disabled',
+    options: []
+  },
   {
     label: '--schedule-low-priority-values-first',
     value: '--schedule-low-priority-values-first',
@@ -225,11 +305,6 @@ const options: BackendParameter[] = [
     value: '--page-size',
     options: []
   },
-  {
-    label: '--hybrid-kvcache-ratio',
-    value: '--hybrid-kvcache-ratio',
-    options: []
-  },
   {
     label: '--swa-full-tokens-ratio',
     value: '--swa-full-tokens-ratio',
@@ -240,22 +315,52 @@ const options: BackendParameter[] = [
     value: '--disable-hybrid-swa-memory',
     options: []
   },
+  {
+    label: '--radix-eviction-policy',
+    value: '--radix-eviction-policy',
+    options: ['lru', 'lfu']
+  },
+  {
+    label: '--enable-prefill-delayer',
+    value: '--enable-prefill-delayer',
+    options: []
+  },
+  {
+    label: '--prefill-delayer-max-delay-passes',
+    value: '--prefill-delayer-max-delay-passes',
+    options: []
+  },
+  {
+    label: '--prefill-delayer-token-usage-low-watermark',
+    value: '--prefill-delayer-token-usage-low-watermark',
+    options: []
+  },
+  {
+    label: '--prefill-delayer-queue-min-ratio',
+    value: '--prefill-delayer-queue-min-ratio',
+    options: []
+  },
+  {
+    label: '--prefill-delayer-max-delay-ms',
+    value: '--prefill-delayer-max-delay-ms',
+    options: []
+  },
+  {
+    label: '--prefill-delayer-forward-passes-buckets',
+    value: '--prefill-delayer-forward-passes-buckets',
+    options: []
+  },
+  {
+    label: '--prefill-delayer-wait-seconds-buckets',
+    value: '--prefill-delayer-wait-seconds-buckets',
+    options: []
+  },
   // Runtime options
   {
     label: '--device',
     value: '--device',
     options: []
   },
-  {
-    label: '--elastic-ep-backend',
-    value: '--elastic-ep-backend',
-    options: []
-  },
-  {
-    label: '--mooncake-ib-device',
-    value: '--mooncake-ib-device',
-    options: []
-  },
   {
     label: '--tensor-parallel-size',
     value: '--tensor-parallel-size',
@@ -276,19 +381,44 @@ const options: BackendParameter[] = [
     value: '--pp-size',
     options: []
   },
+  {
+    label: '--attention-context-parallel-size',
+    value: '--attention-context-parallel-size',
+    options: []
+  },
+  {
+    label: '--attn-cp-size',
+    value: '--attn-cp-size',
+    options: []
+  },
+  {
+    label: '--moe-data-parallel-size',
+    value: '--moe-data-parallel-size',
+    options: []
+  },
+  {
+    label: '--moe-dp-size',
+    value: '--moe-dp-size',
+    options: []
+  },
   {
     label: '--pp-max-micro-batch-size',
     value: '--pp-max-micro-batch-size',
     options: []
   },
+  {
+    label: '--pp-async-batch-depth',
+    value: '--pp-async-batch-depth',
+    options: []
+  },
   {
     label: '--stream-interval',
     value: '--stream-interval',
     options: []
   },
   {
-    label: '--stream-output',
-    value: '--stream-output',
+    label: '--incremental-streaming-output',
+    value: '--incremental-streaming-output',
     options: []
   },
   {
@@ -311,6 +441,11 @@ const options: BackendParameter[] = [
     value: '--watchdog-timeout',
     options: []
   },
+  {
+    label: '--soft-watchdog-timeout',
+    value: '--soft-watchdog-timeout',
+    options: []
+  },
   {
     label: '--dist-timeout',
     value: '--dist-timeout',
@@ -321,6 +456,11 @@ const options: BackendParameter[] = [
     value: '--download-dir',
     options: []
   },
+  {
+    label: '--model-checksum',
+    value: '--model-checksum',
+    options: []
+  },
   {
     label: '--base-gpu-id',
     value: '--base-gpu-id',
@@ -336,6 +476,11 @@ const options: BackendParameter[] = [
     value: '--sleep-on-idle',
     options: []
   },
+  {
+    label: '--custom-sigquit-handler',
+    value: '--custom-sigquit-handler',
+    options: []
+  },
   // Logging
   {
     label: '--log-level',
@@ -358,13 +503,23 @@ const options: BackendParameter[] = [
     options: ['0', '1', '2', '3']
   },
   {
-    label: '--crash-dump-folder',
-    value: '--crash-dump-folder',
+    label: '--log-requests-format',
+    value: '--log-requests-format',
+    options: ['text', 'json']
+  },
+  {
+    label: '--log-requests-target',
+    value: '--log-requests-target',
     options: []
   },
   {
-    label: '--crash-on-nan',
-    value: '--crash-on-nan',
+    label: '--uvicorn-access-log-exclude-prefixes',
+    value: '--uvicorn-access-log-exclude-prefixes',
+    options: []
+  },
+  {
+    label: '--crash-dump-folder',
+    value: '--crash-dump-folder',
     options: []
   },
   {
@@ -377,6 +532,11 @@ const options: BackendParameter[] = [
     value: '--enable-metrics',
     options: []
   },
+  {
+    label: '--enable-mfu-metrics',
+    value: '--enable-mfu-metrics',
+    options: []
+  },
   {
     label: '--enable-metrics-for-all-schedulers',
     value: '--enable-metrics-for-all-schedulers',
@@ -448,8 +608,19 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--oltp-traces-endpoint',
-    value: '--oltp-traces-endpoint',
+    label: '--otlp-traces-endpoint',
+    value: '--otlp-traces-endpoint',
+    options: []
+  },
+  // RequestMetricsExporter configuration
+  {
+    label: '--export-metrics-to-file',
+    value: '--export-metrics-to-file',
+    options: []
+  },
+  {
+    label: '--export-metrics-to-file-dir',
+    value: '--export-metrics-to-file-dir',
     options: []
   },
   // API related
@@ -458,6 +629,11 @@ const options: BackendParameter[] = [
     value: '--api-key',
     options: []
   },
+  {
+    label: '--admin-api-key',
+    value: '--admin-api-key',
+    options: []
+  },
   {
     label: '--served-model-name',
     value: '--served-model-name',
@@ -473,6 +649,11 @@ const options: BackendParameter[] = [
     value: '--chat-template',
     options: []
   },
+  {
+    label: '--hf-chat-template-name',
+    value: '--hf-chat-template-name',
+    options: []
+  },
   {
     label: '--completion-template',
     value: '--completion-template',
@@ -510,6 +691,7 @@ const options: BackendParameter[] = [
       'deepseekv31',
       'glm',
       'glm45',
+      'glm47',
       'gpt-oss',
       'kimi_k2',
       'llama3',
@@ -518,25 +700,26 @@ const options: BackendParameter[] = [
       'qwen',
       'qwen25',
       'qwen3_coder',
-      'step3'
+      'step3',
+      'gigachat3'
     ]
   },
-  {
-    label: '--sampling-defaults',
-    value: '--sampling-defaults',
-    options: ['openai', 'model']
-  },
   {
     label: '--tool-server',
     value: '--tool-server',
     options: []
   },
+  {
+    label: '--sampling-defaults',
+    value: '--sampling-defaults',
+    options: ['openai', 'model']
+  },
+  // Data parallelism
   {
     label: '--data-parallel-size',
     value: '--data-parallel-size',
     options: []
   },
-  // Data parallelism
   {
     label: '--dp-size',
     value: '--dp-size',
@@ -545,17 +728,13 @@ const options: BackendParameter[] = [
   {
     label: '--load-balance-method',
     value: '--load-balance-method',
-    options: ['round_robin', 'shortest_queue', 'minimum_tokens']
-  },
-  {
-    label: '--load-watch-interval',
-    value: '--load-watch-interval',
-    options: []
-  },
-  {
-    label: '--prefill-round-robin-balance',
-    value: '--prefill-round-robin-balance',
-    options: []
+    options: [
+      'auto',
+      'round_robin',
+      'follow_bootstrap_room',
+      'total_requests',
+      'total_tokens'
+    ]
   },
   // Multi-node distributed serving
   {
@@ -595,6 +774,11 @@ const options: BackendParameter[] = [
     value: '--enable-lora',
     options: []
   },
+  {
+    label: '--enable-lora-overlap-loading',
+    value: '--enable-lora-overlap-loading',
+    options: []
+  },
   {
     label: '--max-lora-rank',
     value: '--max-lora-rank',
@@ -619,7 +803,7 @@ const options: BackendParameter[] = [
   {
     label: '--lora-paths',
     value: '--lora-paths',
-    options: ['{"lora_name": str, "lora_path": str, "pinned": bool}']
+    options: []
   },
   {
     label: '--max-loras-per-batch',
@@ -639,14 +823,19 @@ const options: BackendParameter[] = [
   {
     label: '--lora-backend',
     value: '--lora-backend',
-    options: ['triton', 'csgmv']
+    options: ['triton', 'csgmv', 'ascend', 'torch_native']
   },
   {
     label: '--max-lora-chunk-size',
     value: '--max-lora-chunk-size',
     options: ['16', '32', '64', '128']
   },
-  // Kernel backend
+  {
+    label: '--lora-drain-wait-threshold',
+    value: '--lora-drain-wait-threshold',
+    options: []
+  },
+  // Kernel Backends (Attention, Sampling, Grammar, GEMM)
   {
     label: '--attention-backend',
     value: '--attention-backend',
@@ -654,6 +843,7 @@ const options: BackendParameter[] = [
       'triton',
       'torch_native',
       'flex_attention',
+      'dsa',
       'nsa',
       'cutlass_mla',
       'fa3',
@@ -676,6 +866,7 @@ const options: BackendParameter[] = [
       'triton',
       'torch_native',
       'flex_attention',
+      'dsa',
       'nsa',
       'cutlass_mla',
       'fa3',
@@ -698,6 +889,7 @@ const options: BackendParameter[] = [
       'triton',
       'torch_native',
       'flex_attention',
+      'dsa',
       'nsa',
       'cutlass_mla',
       'fa3',
@@ -716,7 +908,7 @@ const options: BackendParameter[] = [
   {
     label: '--sampling-backend',
     value: '--sampling-backend',
-    options: ['flashinfer', 'pytorch']
+    options: ['flashinfer', 'pytorch', 'ascend']
   },
   {
     label: '--grammar-backend',
@@ -726,17 +918,66 @@ const options: BackendParameter[] = [
   {
     label: '--mm-attention-backend',
     value: '--mm-attention-backend',
-    options: ['sdpa', 'fa3', 'triton_attn', 'ascend_attn']
+    options: ['sdpa', 'fa3', 'fa4', 'triton_attn', 'ascend_attn', 'aiter_attn']
   },
   {
-    label: '--nsa-prefill',
-    value: '--nsa-prefill',
-    options: ['flashmla_sparse', 'flashmla_decode', 'fa3', 'tilelang', 'aiter']
+    label: '--dsa-prefill-backend',
+    value: '--dsa-prefill-backend',
+    options: [
+      'flashmla_sparse',
+      'flashmla_kv',
+      'flashmla_auto',
+      'fa3',
+      'tilelang',
+      'aiter',
+      'trtllm'
+    ]
   },
   {
-    label: '--nsa-decode',
-    value: '--nsa-decode',
-    options: ['flashmla_prefill', 'flashmla_kv', 'fa3', 'tilelang', 'aiter']
+    label: '--dsa-decode-backend',
+    value: '--dsa-decode-backend',
+    options: [
+      'flashmla_sparse',
+      'flashmla_kv',
+      'fa3',
+      'tilelang',
+      'aiter',
+      'trtllm'
+    ]
+  },
+  {
+    label: '--dsa-topk-backend',
+    value: '--dsa-topk-backend',
+    options: ['sgl-kernel', 'torch', 'flashinfer']
+  },
+  {
+    label: '--fp8-gemm-backend',
+    value: '--fp8-gemm-backend',
+    options: [
+      'auto',
+      'deep_gemm',
+      'flashinfer_trtllm',
+      'flashinfer_cutlass',
+      'flashinfer_deepgemm',
+      'cutlass',
+      'triton',
+      'aiter'
+    ]
+  },
+  {
+    label: '--fp4-gemm-backend',
+    value: '--fp4-gemm-backend',
+    options: [
+      'auto',
+      'flashinfer_cudnn',
+      'flashinfer_cutlass',
+      'flashinfer_trtllm'
+    ]
+  },
+  {
+    label: '--disable-flashinfer-autotune',
+    value: '--disable-flashinfer-autotune',
+    options: []
   },
   // Speculative decoding
   {
@@ -759,6 +1000,11 @@ const options: BackendParameter[] = [
     value: '--speculative-draft-model-revision',
     options: []
   },
+  {
+    label: '--speculative-draft-load-format',
+    value: '--speculative-draft-load-format',
+    options: []
+  },
   {
     label: '--speculative-num-steps',
     value: '--speculative-num-steps',
@@ -794,17 +1040,27 @@ const options: BackendParameter[] = [
     value: '--speculative-attention-mode',
     options: ['prefill', 'decode']
   },
+  {
+    label: '--speculative-draft-attention-backend',
+    value: '--speculative-draft-attention-backend',
+    options: []
+  },
+  {
+    label: '--speculative-moe-runner-backend',
+    value: '--speculative-moe-runner-backend',
+    options: []
+  },
+  {
+    label: '--speculative-moe-a2a-backend',
+    value: '--speculative-moe-a2a-backend',
+    options: []
+  },
+  {
+    label: '--speculative-draft-model-quantization',
+    value: '--speculative-draft-model-quantization',
+    options: []
+  },
   // Ngram speculative decoding
-  {
-    label: '--speculative-ngram-min-match-window-size',
-    value: '--speculative-ngram-min-match-window-size',
-    options: []
-  },
-  {
-    label: '--speculative-ngram-max-match-window-size',
-    value: '--speculative-ngram-max-match-window-size',
-    options: []
-  },
   {
     label: '--speculative-ngram-min-bfs-breadth',
     value: '--speculative-ngram-min-bfs-breadth',
@@ -821,8 +1077,8 @@ const options: BackendParameter[] = [
     options: ['BFS', 'PROB']
   },
   {
-    label: '--speculative-ngram-branch-length',
-    value: '--speculative-ngram-branch-length',
+    label: '--speculative-ngram-max-trie-depth',
+    value: '--speculative-ngram-max-trie-depth',
     options: []
   },
   {
@@ -830,7 +1086,13 @@ const options: BackendParameter[] = [
     value: '--speculative-ngram-capacity',
     options: []
   },
-  // Expert parallelism
+  // Multi-layer Eagle speculative decoding
+  {
+    label: '--enable-multi-layer-eagle',
+    value: '--enable-multi-layer-eagle',
+    options: []
+  },
+  // MoE
   {
     label: '--expert-parallel-size',
     value: '--expert-parallel-size',
@@ -849,7 +1111,7 @@ const options: BackendParameter[] = [
   {
     label: '--moe-a2a-backend',
     value: '--moe-a2a-backend',
-    options: ['none', 'deepep']
+    options: ['none', 'deepep', 'mooncake', 'mori', 'nixl', 'ascend_fuseep']
   },
   {
     label: '--moe-runner-backend',
@@ -860,9 +1122,11 @@ const options: BackendParameter[] = [
       'triton',
       'triton_kernel',
       'flashinfer_trtllm',
+      'flashinfer_trtllm_routed',
       'flashinfer_cutlass',
       'flashinfer_mxfp4',
-      'flashinfer_cutedsl'
+      'flashinfer_cutedsl',
+      'cutlass'
     ]
   },
   {
@@ -875,11 +1139,21 @@ const options: BackendParameter[] = [
     value: '--enable-flashinfer-allreduce-fusion',
     options: []
   },
+  {
+    label: '--enable-aiter-allreduce-fusion',
+    value: '--enable-aiter-allreduce-fusion',
+    options: []
+  },
   {
     label: '--deepep-mode',
     value: '--deepep-mode',
     options: ['normal', 'low_latency', 'auto']
   },
+  {
+    label: '--deepep-dispatcher-output-dtype',
+    value: '--deepep-dispatcher-output-dtype',
+    options: ['bf16', 'fp8', 'int8', 'nvfp4', 'auto']
+  },
   {
     label: '--ep-num-redundant-experts',
     value: '--ep-num-redundant-experts',
@@ -945,6 +1219,31 @@ const options: BackendParameter[] = [
     value: '--moe-dense-tp-size',
     options: []
   },
+  {
+    label: '--elastic-ep-backend',
+    value: '--elastic-ep-backend',
+    options: ['none', 'mooncake']
+  },
+  {
+    label: '--enable-elastic-expert-backup',
+    value: '--enable-elastic-expert-backup',
+    options: []
+  },
+  {
+    label: '--mooncake-ib-device',
+    value: '--mooncake-ib-device',
+    options: []
+  },
+  {
+    label: '--enable-deepep-waterfill',
+    value: '--enable-deepep-waterfill',
+    options: []
+  },
+  {
+    label: '--elastic-ep-rejoin',
+    value: '--elastic-ep-rejoin',
+    options: []
+  },
   // Mamba Cache
   {
     label: '--max-mamba-cache-size',
@@ -954,17 +1253,21 @@ const options: BackendParameter[] = [
   {
     label: '--mamba-ssm-dtype',
     value: '--mamba-ssm-dtype',
-    options: ['float32', 'bfloat16']
+    options: ['float32', 'bfloat16', 'float16']
   },
   {
     label: '--mamba-full-memory-ratio',
     value: '--mamba-full-memory-ratio',
     options: []
   },
-  // Args for multi-item scoring
   {
-    label: '--multi-item-scoring-delimiter',
-    value: '--multi-item-scoring-delimiter',
+    label: '--mamba-scheduler-strategy',
+    value: '--mamba-scheduler-strategy',
+    options: ['auto', 'no_buffer', 'extra_buffer']
+  },
+  {
+    label: '--mamba-track-interval',
+    value: '--mamba-track-interval',
     options: []
   },
   // Hierarchical cache
@@ -988,20 +1291,21 @@ const options: BackendParameter[] = [
     value: '--hicache-write-policy',
     options: ['write_back', 'write_through', 'write_through_selective']
   },
-  {
-    label: '--radix-eviction-policy',
-    value: '--radix-eviction-policy',
-    options: ['lru', 'lfu']
-  },
   {
     label: '--hicache-io-backend',
     value: '--hicache-io-backend',
-    options: ['direct', 'kernel']
+    options: ['direct', 'kernel', 'kernel_ascend']
   },
   {
     label: '--hicache-mem-layout',
     value: '--hicache-mem-layout',
-    options: ['layer_first', 'page_first', 'page_first_direct']
+    options: [
+      'layer_first',
+      'page_first',
+      'page_first_direct',
+      'page_first_kv_split',
+      'page_head'
+    ]
   },
   {
     label: '--hicache-storage-backend',
@@ -1018,43 +1322,18 @@ const options: BackendParameter[] = [
     value: '--hicache-storage-backend-extra-config',
     options: []
   },
+  // Hierarchical sparse attention
+  {
+    label: '--hierarchical-sparse-attention-extra-config',
+    value: '--hierarchical-sparse-attention-extra-config',
+    options: []
+  },
   // LMCache
   {
     label: '--enable-lmcache',
     value: '--enable-lmcache',
     options: []
   },
-  // Double Sparsity
-  {
-    label: '--enable-double-sparsity',
-    value: '--enable-double-sparsity',
-    options: []
-  },
-  {
-    label: '--ds-channel-config-path',
-    value: '--ds-channel-config-path',
-    options: []
-  },
-  {
-    label: '--ds-heavy-channel-num',
-    value: '--ds-heavy-channel-num',
-    options: []
-  },
-  {
-    label: '--ds-heavy-token-num',
-    value: '--ds-heavy-token-num',
-    options: []
-  },
-  {
-    label: '--ds-heavy-channel-type',
-    value: '--ds-heavy-channel-type',
-    options: []
-  },
-  {
-    label: '--ds-sparse-decode-threshold',
-    value: '--ds-sparse-decode-threshold',
-    options: []
-  },
   // Offloading
   {
     label: '--cpu-offload-gb',
@@ -1081,6 +1360,12 @@ const options: BackendParameter[] = [
     value: '--offload-mode',
     options: []
   },
+  // Args for multi-item scoring
+  {
+    label: '--multi-item-scoring-delimiter',
+    value: '--multi-item-scoring-delimiter',
+    options: []
+  },
   // Optimization/debug options
   {
     label: '--disable-radix-cache',
@@ -1117,6 +1402,11 @@ const options: BackendParameter[] = [
     value: '--enable-cudagraph-gc',
     options: []
   },
+  {
+    label: '--enable-layerwise-nvtx-marker',
+    value: '--enable-layerwise-nvtx-marker',
+    options: []
+  },
   {
     label: '--enable-nccl-nvls',
     value: '--enable-nccl-nvls',
@@ -1137,6 +1427,11 @@ const options: BackendParameter[] = [
     value: '--enable-tokenizer-batch-encode',
     options: []
   },
+  {
+    label: '--disable-tokenizer-batch-decode',
+    value: '--disable-tokenizer-batch-decode',
+    options: []
+  },
   {
     label: '--disable-outlines-disk-cache',
     value: '--disable-outlines-disk-cache',
@@ -1198,8 +1493,18 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--enable-piecewise-cuda-graph',
-    value: '--enable-piecewise-cuda-graph',
+    label: '--enable-torch-compile-debug-mode',
+    value: '--enable-torch-compile-debug-mode',
+    options: []
+  },
+  {
+    label: '--disable-piecewise-cuda-graph',
+    value: '--disable-piecewise-cuda-graph',
+    options: []
+  },
+  {
+    label: '--enforce-piecewise-cuda-graph',
+    value: '--enforce-piecewise-cuda-graph',
     options: []
   },
   {
@@ -1207,6 +1512,11 @@ const options: BackendParameter[] = [
     value: '--piecewise-cuda-graph-tokens',
     options: []
   },
+  {
+    label: '--piecewise-cuda-graph-compiler',
+    value: '--piecewise-cuda-graph-compiler',
+    options: ['eager', 'inductor']
+  },
   {
     label: '--torch-compile-max-bs',
     value: '--torch-compile-max-bs',
@@ -1267,6 +1577,11 @@ const options: BackendParameter[] = [
     value: '--enable-weights-cpu-backup',
     options: []
   },
+  {
+    label: '--enable-draft-weights-cpu-backup',
+    value: '--enable-draft-weights-cpu-backup',
+    options: []
+  },
   {
     label: '--allow-auto-truncate',
     value: '--allow-auto-truncate',
@@ -1307,6 +1622,11 @@ const options: BackendParameter[] = [
     value: '--enable-return-hidden-states',
     options: []
   },
+  {
+    label: '--enable-return-routed-experts',
+    value: '--enable-return-routed-experts',
+    options: []
+  },
   {
     label: '--scheduler-recv-interval',
     value: '--scheduler-recv-interval',
@@ -1317,22 +1637,42 @@ const options: BackendParameter[] = [
     value: '--numa-node',
     options: []
   },
-  // Debug tensor dumps
   {
-    label: '--debug-tensor-dump-output-folder',
-    value: '--debug-tensor-dump-output-folder',
+    label: '--enable-deterministic-inference',
+    value: '--enable-deterministic-inference',
     options: []
   },
   {
-    label: '--debug-tensor-dump-input-file',
-    value: '--debug-tensor-dump-input-file',
+    label: '--rl-on-policy-target',
+    value: '--rl-on-policy-target',
+    options: ['fsdp']
+  },
+  {
+    label: '--enable-attn-tp-input-scattered',
+    value: '--enable-attn-tp-input-scattered',
     options: []
   },
   {
-    label: '--debug-tensor-dump-inject',
-    value: '--debug-tensor-dump-inject',
+    label: '--enable-dsa-prefill-context-parallel',
+    value: '--enable-dsa-prefill-context-parallel',
     options: []
   },
+  {
+    label: '--dsa-prefill-cp-mode',
+    value: '--dsa-prefill-cp-mode',
+    options: ['in-seq-split', 'round-robin-split']
+  },
+  {
+    label: '--enable-fused-qk-norm-rope',
+    value: '--enable-fused-qk-norm-rope',
+    options: []
+  },
+  {
+    label: '--enable-precise-embedding-interpolation',
+    value: '--enable-precise-embedding-interpolation',
+    options: []
+  },
+  // Dynamic batch tokenizer
   {
     label: '--enable-dynamic-batch-tokenizer',
     value: '--enable-dynamic-batch-tokenizer',
@@ -1348,6 +1688,27 @@ const options: BackendParameter[] = [
     value: '--dynamic-batch-tokenizer-batch-timeout',
     options: []
   },
+  // Debug tensor dumps
+  {
+    label: '--debug-tensor-dump-output-folder',
+    value: '--debug-tensor-dump-output-folder',
+    options: []
+  },
+  {
+    label: '--debug-tensor-dump-layers',
+    value: '--debug-tensor-dump-layers',
+    options: []
+  },
+  {
+    label: '--debug-tensor-dump-input-file',
+    value: '--debug-tensor-dump-input-file',
+    options: []
+  },
+  {
+    label: '--debug-tensor-dump-inject',
+    value: '--debug-tensor-dump-inject',
+    options: []
+  },
   // PD disaggregation
   {
     label: '--disaggregation-mode',
@@ -1364,21 +1725,6 @@ const options: BackendParameter[] = [
     value: '--disaggregation-bootstrap-port',
     options: []
   },
-  {
-    label: '--disaggregation-decode-tp',
-    value: '--disaggregation-decode-tp',
-    options: []
-  },
-  {
-    label: '--disaggregation-decode-dp',
-    value: '--disaggregation-decode-dp',
-    options: []
-  },
-  {
-    label: '--disaggregation-prefill-pp',
-    value: '--disaggregation-prefill-pp',
-    options: []
-  },
   {
     label: '--disaggregation-ib-device',
     value: '--disaggregation-ib-device',
@@ -1399,6 +1745,27 @@ const options: BackendParameter[] = [
     value: '--disaggregation-decode-polling-interval',
     options: []
   },
+  // Encode prefill disaggregation
+  {
+    label: '--encoder-only',
+    value: '--encoder-only',
+    options: []
+  },
+  {
+    label: '--language-only',
+    value: '--language-only',
+    options: []
+  },
+  {
+    label: '--encoder-transfer-backend',
+    value: '--encoder-transfer-backend',
+    options: ['zmq_to_scheduler', 'zmq_to_tokenizer', 'mooncake']
+  },
+  {
+    label: '--encoder-urls',
+    value: '--encoder-urls',
+    options: []
+  },
   // Custom weight loader
   {
     label: '--custom-weight-loader',
@@ -1410,6 +1777,16 @@ const options: BackendParameter[] = [
     value: '--weight-loader-disable-mmap',
     options: []
   },
+  {
+    label: '--weight-loader-prefetch-checkpoints',
+    value: '--weight-loader-prefetch-checkpoints',
+    options: []
+  },
+  {
+    label: '--weight-loader-prefetch-num-threads',
+    value: '--weight-loader-prefetch-num-threads',
+    options: []
+  },
   {
     label: '--remote-instance-weight-loader-seed-instance-ip',
     value: '--remote-instance-weight-loader-seed-instance-ip',
@@ -1425,6 +1802,16 @@ const options: BackendParameter[] = [
     value: '--remote-instance-weight-loader-send-weights-group-ports',
     options: []
   },
+  {
+    label: '--remote-instance-weight-loader-backend',
+    value: '--remote-instance-weight-loader-backend',
+    options: ['transfer_engine', 'nccl']
+  },
+  {
+    label: '--remote-instance-weight-loader-start-seed-via-transfer-engine',
+    value: '--remote-instance-weight-loader-start-seed-via-transfer-engine',
+    options: []
+  },
   // For PD-Multiplexing
   {
     label: '--enable-pdmux',
@@ -1441,17 +1828,63 @@ const options: BackendParameter[] = [
     value: '--sm-group-num',
     options: []
   },
-  // For deterministic inference
-  {
-    label: '--enable-deterministic-inference',
-    value: '--enable-deterministic-inference',
-    options: []
-  },
   // Configuration file support
   {
     label: '--config',
     value: '--config',
     options: []
+  },
+  // For Multi-Modal
+  {
+    label: '--mm-max-concurrent-calls',
+    value: '--mm-max-concurrent-calls',
+    options: []
+  },
+  {
+    label: '--mm-per-request-timeout',
+    value: '--mm-per-request-timeout',
+    options: []
+  },
+  {
+    label: '--enable-broadcast-mm-inputs-process',
+    value: '--enable-broadcast-mm-inputs-process',
+    options: []
+  },
+  {
+    label: '--mm-process-config',
+    value: '--mm-process-config',
+    options: []
+  },
+  {
+    label: '--mm-enable-dp-encoder',
+    value: '--mm-enable-dp-encoder',
+    options: []
+  },
+  {
+    label: '--limit-mm-data-per-request',
+    value: '--limit-mm-data-per-request',
+    options: []
+  },
+  {
+    label: '--enable-mm-global-cache',
+    value: '--enable-mm-global-cache',
+    options: []
+  },
+  // For checkpoint decryption
+  {
+    label: '--decrypted-config-file',
+    value: '--decrypted-config-file',
+    options: []
+  },
+  {
+    label: '--decrypted-draft-config-file',
+    value: '--decrypted-draft-config-file',
+    options: []
+  },
+  {
+    label: '--enable-prefix-mm-cache',
+    value: '--enable-prefix-mm-cache',
+    options: []
   }
 ];
 
diff --git a/src/pages/llmodels/constants/backend-parameters/vllm.ts b/src/pages/llmodels/constants/backend-parameters/vllm.ts
index a5b08dc6..5397f097 100644
--- a/src/pages/llmodels/constants/backend-parameters/vllm.ts
+++ b/src/pages/llmodels/constants/backend-parameters/vllm.ts
@@ -1,16 +1,234 @@
 import { BackendParameter } from './index';
 
+// Generated from the vLLM `vllm serve` CLI reference:
+// https://docs.vllm.ai/en/latest/cli/serve/
+// Grouped by the config sections in the docs. Long-form flags only
+// (short aliases and `--no-*` negations are omitted). `options` lists the
+// fixed choice values when the flag accepts a restricted set.
 const options: BackendParameter[] = [
+  // General
+  {
+    label: '--headless',
+    value: '--headless',
+    options: []
+  },
+  {
+    label: '--api-server-count',
+    value: '--api-server-count',
+    options: []
+  },
+  {
+    label: '--config',
+    value: '--config',
+    options: []
+  },
+  {
+    label: '--grpc',
+    value: '--grpc',
+    options: []
+  },
+  {
+    label: '--disable-log-stats',
+    value: '--disable-log-stats',
+    options: []
+  },
+  {
+    label: '--aggregate-engine-logging',
+    value: '--aggregate-engine-logging',
+    options: []
+  },
+  {
+    label: '--fail-on-environ-validation',
+    value: '--fail-on-environ-validation',
+    options: []
+  },
+  {
+    label: '--shutdown-timeout',
+    value: '--shutdown-timeout',
+    options: []
+  },
+  {
+    label: '--gdn-prefill-backend',
+    value: '--gdn-prefill-backend',
+    options: ['flashinfer', 'triton', 'cutedsl']
+  },
+  {
+    label: '--enable-log-requests',
+    value: '--enable-log-requests',
+    options: []
+  },
+
+  // Frontend
+  {
+    label: '--lora-modules',
+    value: '--lora-modules',
+    options: []
+  },
+  {
+    label: '--chat-template',
+    value: '--chat-template',
+    options: []
+  },
+  {
+    label: '--chat-template-content-format',
+    value: '--chat-template-content-format',
+    options: ['auto', 'openai', 'string']
+  },
+  {
+    label: '--trust-request-chat-template',
+    value: '--trust-request-chat-template',
+    options: []
+  },
+  {
+    label: '--default-chat-template-kwargs',
+    value: '--default-chat-template-kwargs',
+    options: []
+  },
+  {
+    label: '--response-role',
+    value: '--response-role',
+    options: []
+  },
+  {
+    label: '--return-tokens-as-token-ids',
+    value: '--return-tokens-as-token-ids',
+    options: []
+  },
+  {
+    label: '--enable-auto-tool-choice',
+    value: '--enable-auto-tool-choice',
+    options: []
+  },
+  {
+    label: '--exclude-tools-when-tool-choice-none',
+    value: '--exclude-tools-when-tool-choice-none',
+    options: []
+  },
+  {
+    label: '--tool-call-parser',
+    value: '--tool-call-parser',
+    options: []
+  },
+  {
+    label: '--tool-parser-plugin',
+    value: '--tool-parser-plugin',
+    options: []
+  },
+  {
+    label: '--tool-server',
+    value: '--tool-server',
+    options: []
+  },
+  {
+    label: '--log-config-file',
+    value: '--log-config-file',
+    options: []
+  },
+  {
+    label: '--max-log-len',
+    value: '--max-log-len',
+    options: []
+  },
+  {
+    label: '--enable-prompt-tokens-details',
+    value: '--enable-prompt-tokens-details',
+    options: []
+  },
+  {
+    label: '--enable-server-load-tracking',
+    value: '--enable-server-load-tracking',
+    options: []
+  },
+  {
+    label: '--enable-force-include-usage',
+    value: '--enable-force-include-usage',
+    options: []
+  },
+  {
+    label: '--enable-tokenizer-info-endpoint',
+    value: '--enable-tokenizer-info-endpoint',
+    options: []
+  },
+  {
+    label: '--enable-log-outputs',
+    value: '--enable-log-outputs',
+    options: []
+  },
+  {
+    label: '--enable-log-deltas',
+    value: '--enable-log-deltas',
+    options: []
+  },
+  {
+    label: '--log-error-stack',
+    value: '--log-error-stack',
+    options: []
+  },
+  {
+    label: '--tokens-only',
+    value: '--tokens-only',
+    options: []
+  },
+  {
+    label: '--fingerprint-mode',
+    value: '--fingerprint-mode',
+    options: ['custom', 'full', 'hash', 'none']
+  },
+  {
+    label: '--fingerprint-value',
+    value: '--fingerprint-value',
+    options: []
+  },
+  {
+    label: '--host',
+    value: '--host',
+    options: []
+  },
+  {
+    label: '--port',
+    value: '--port',
+    options: []
+  },
+  {
+    label: '--data-parallel-supervisor-port',
+    value: '--data-parallel-supervisor-port',
+    options: []
+  },
+  {
+    label: '--dp-supervisor-probe-interval-s',
+    value: '--dp-supervisor-probe-interval-s',
+    options: []
+  },
+  {
+    label: '--dp-supervisor-probe-timeout-s',
+    value: '--dp-supervisor-probe-timeout-s',
+    options: []
+  },
+  {
+    label: '--dp-supervisor-probe-failure-threshold',
+    value: '--dp-supervisor-probe-failure-threshold',
+    options: []
+  },
+  {
+    label: '--uds',
+    value: '--uds',
+    options: []
+  },
   {
     label: '--uvicorn-log-level',
     value: '--uvicorn-log-level',
-    options: ['debug', 'info', 'warning', 'error', 'critical', 'trace']
+    options: ['critical', 'debug', 'error', 'info', 'trace', 'warning']
   },
   {
     label: '--disable-uvicorn-access-log',
     value: '--disable-uvicorn-access-log',
     options: []
   },
+  {
+    label: '--disable-access-log-for-endpoints',
+    value: '--disable-access-log-for-endpoints',
+    options: []
+  },
   {
     label: '--allow-credentials',
     value: '--allow-credentials',
@@ -36,26 +254,6 @@ const options: BackendParameter[] = [
     value: '--api-key',
     options: []
   },
-  {
-    label: '--lora-modules',
-    value: '--lora-modules',
-    options: []
-  },
-  {
-    label: '--prompt-adapters',
-    value: '--prompt-adapters',
-    options: []
-  },
-  {
-    label: '--chat-template',
-    value: '--chat-template',
-    options: []
-  },
-  {
-    label: '--response-role',
-    value: '--response-role',
-    options: []
-  },
   {
     label: '--ssl-keyfile',
     value: '--ssl-keyfile',
@@ -67,8 +265,8 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--data-parallel-size-local',
-    value: '--data-parallel-size-local',
+    label: '--ssl-ca-certs',
+    value: '--ssl-ca-certs',
     options: []
   },
   {
@@ -77,13 +275,13 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--ssl-ca-certs',
-    value: '--ssl-ca-certs',
+    label: '--ssl-cert-reqs',
+    value: '--ssl-cert-reqs',
     options: []
   },
   {
-    label: '--ssl-cert-reqs',
-    value: '--ssl-cert-reqs',
+    label: '--ssl-ciphers',
+    value: '--ssl-ciphers',
     options: []
   },
   {
@@ -96,93 +294,91 @@ const options: BackendParameter[] = [
     value: '--middleware',
     options: []
   },
-  {
-    label: '--return-tokens-as-token-ids',
-    value: '--return-tokens-as-token-ids',
-    options: []
-  },
-  {
-    label: '--disable-frontend-multiprocessing',
-    value: '--disable-frontend-multiprocessing',
-    options: []
-  },
   {
     label: '--enable-request-id-headers',
     value: '--enable-request-id-headers',
     options: []
   },
   {
-    label: '--enable-auto-tool-choice',
-    value: '--enable-auto-tool-choice',
+    label: '--disable-fastapi-docs',
+    value: '--disable-fastapi-docs',
     options: []
   },
   {
-    label: '--tool-call-parser',
-    value: '--tool-call-parser',
-    options: [
-      'phi4_mini_json',
-      'llama3_json',
-      'llama4_json',
-      'pythonic',
-      'jamba',
-      'xlam',
-      'llama4_pythonic',
-      'hunyuan_a13b',
-      'mistral',
-      'deepseek_v3',
-      'kimi_k2',
-      'step3',
-      'qwen3_coder',
-      'hermes',
-      'glm45',
-      'granite',
-      'minimax',
-      'granite-20b-fc',
-      'internlm'
-    ]
-  },
-  {
-    label: '--task',
-    value: '--task',
-    options: [
-      'auto',
-      'generate',
-      'embedding',
-      'embed',
-      'classify',
-      'score',
-      'reward',
-      'transcription'
-    ]
-  },
-  {
-    label: '--allowed-local-media-path',
-    value: '--allowed-local-media-path',
+    label: '--h11-max-incomplete-event-size',
+    value: '--h11-max-incomplete-event-size',
     options: []
   },
   {
-    label: '--tool-parser-plugin',
-    value: '--tool-parser-plugin',
+    label: '--h11-max-header-count',
+    value: '--h11-max-header-count',
     options: []
   },
+  {
+    label: '--enable-offline-docs',
+    value: '--enable-offline-docs',
+    options: []
+  },
+  {
+    label: '--enable-flash-late-interaction',
+    value: '--enable-flash-late-interaction',
+    options: []
+  },
+
+  // ModelConfig
   {
     label: '--model',
     value: '--model',
     options: []
   },
+  {
+    label: '--runner',
+    value: '--runner',
+    options: ['auto', 'draft', 'generate', 'pooling']
+  },
+  {
+    label: '--convert',
+    value: '--convert',
+    options: ['auto', 'classify', 'embed', 'none']
+  },
   {
     label: '--tokenizer',
     value: '--tokenizer',
     options: []
   },
+  {
+    label: '--tokenizer-mode',
+    value: '--tokenizer-mode',
+    options: ['auto', 'deepseek_v32', 'deepseek_v4', 'hf', 'mistral', 'slow']
+  },
+  {
+    label: '--trust-remote-code',
+    value: '--trust-remote-code',
+    options: []
+  },
+  {
+    label: '--dtype',
+    value: '--dtype',
+    options: ['auto', 'bfloat16', 'float', 'float16', 'float32', 'half']
+  },
+  {
+    label: '--seed',
+    value: '--seed',
+    options: []
+  },
   {
     label: '--hf-config-path',
     value: '--hf-config-path',
     options: []
   },
   {
-    label: '--skip-tokenizer-init',
-    value: '--skip-tokenizer-init',
+    label: '--allowed-local-media-path',
+    value: '--allowed-local-media-path',
+    options: []
+  },
+  {
+    label: '--allowed-media-domains',
+    value: '--allowed-media-domains',
     options: []
   },
   {
@@ -201,66 +397,59 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--tokenizer-mode',
-    value: '--tokenizer-mode',
-    options: ['auto', 'slow', 'mistral', 'custom']
-  },
-  {
-    label: '--trust-remote-code',
-    value: '--trust-remote-code',
+    label: '--max-model-len',
+    value: '--max-model-len',
     options: []
   },
   {
-    label: '--download-dir',
-    value: '--download-dir',
+    label: '--quantization',
+    value: '--quantization',
     options: []
   },
   {
-    label: '--load-format',
-    value: '--load-format',
+    label: '--quantization-config',
+    value: '--quantization-config',
+    options: []
+  },
+  {
+    label: '--allow-deprecated-quantization',
+    value: '--allow-deprecated-quantization',
+    options: []
+  },
+  {
+    label: '--enforce-eager',
+    value: '--enforce-eager',
+    options: []
+  },
+  {
+    label: '--enable-return-routed-experts',
+    value: '--enable-return-routed-experts',
+    options: []
+  },
+  {
+    label: '--max-logprobs',
+    value: '--max-logprobs',
+    options: []
+  },
+  {
+    label: '--logprobs-mode',
+    value: '--logprobs-mode',
     options: [
-      'auto',
-      'pt',
-      'safetensors',
-      'npcache',
-      'dummy',
-      'tensorizer',
-      'sharded_state',
-      'gguf',
-      'bitsandbytes',
-      'mistral',
-      'runai_streamer',
-      'fastsafetensors'
+      'processed_logits',
+      'processed_logprobs',
+      'raw_logits',
+      'raw_logprobs'
     ]
   },
   {
-    label: '--chat-template-content-format',
-    value: '--chat-template-content-format',
-    options: ['auto', 'string', 'openai']
-  },
-  {
-    label: '--enable-reasoning',
-    value: '--enable-reasoning',
+    label: '--use-fp64-gumbel',
+    value: '--use-fp64-gumbel',
     options: []
   },
   {
-    label: '--reasoning-parser',
-    value: '--reasoning-parser',
-    options: [
-      'deepseek_r1',
-      'glm45',
-      'GptOss',
-      'granite',
-      'hunyuan_a13b',
-      'mistral',
-      'qwen3',
-      'step3'
-    ]
-  },
-  {
-    label: '--config-format',
-    value: '--config-format',
-    options: ['auto', 'hf', 'mistral']
+    label: '--disable-sliding-window',
+    value: '--disable-sliding-window',
+    options: []
   },
   {
     label: '--disable-cascade-attn',
@@ -268,59 +457,267 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--dtype',
-    value: '--dtype',
-    options: ['auto', 'half', 'float16', 'bfloat16', 'float', 'float32']
-  },
-  {
-    label: '--kv-cache-dtype',
-    value: '--kv-cache-dtype',
-    options: ['auto', 'fp8', 'fp8_e5m2', 'fp8_e4m3']
-  },
-  {
-    label: '--disable-chunked-mm-input',
-    value: '--disable-chunked-mm-input',
+    label: '--skip-tokenizer-init',
+    value: '--skip-tokenizer-init',
     options: []
   },
   {
-    label: 'DISABLE_CHUNKED_MM_INPUT',
-    value: 'DISABLE_CHUNKED_MM_INPUT',
+    label: '--enable-prompt-embeds',
+    value: '--enable-prompt-embeds',
     options: []
   },
   {
-    label: '--quantization-param-path',
-    value: '--quantization-param-path',
+    label: '--served-model-name',
+    value: '--served-model-name',
     options: []
   },
   {
-    label: '--max-model-len',
-    value: '--max-model-len',
+    label: '--config-format',
+    value: '--config-format',
+    options: ['auto', 'hf', 'mistral']
+  },
+  {
+    label: '--hf-token',
+    value: '--hf-token',
     options: []
   },
   {
-    label: '--guided-decoding-backend',
-    value: '--guided-decoding-backend',
-    options: ['outlines', 'lm-format-enforcer', 'xgrammar']
+    label: '--hf-overrides',
+    value: '--hf-overrides',
+    options: []
   },
   {
-    label: '--logits-processor-pattern',
-    value: '--logits-processor-pattern',
+    label: '--pooler-config',
+    value: '--pooler-config',
+    options: []
+  },
+  {
+    label: '--generation-config',
+    value: '--generation-config',
+    options: []
+  },
+  {
+    label: '--override-generation-config',
+    value: '--override-generation-config',
+    options: []
+  },
+  {
+    label: '--enable-sleep-mode',
+    value: '--enable-sleep-mode',
+    options: []
+  },
+  {
+    label: '--enable-cumem-allocator',
+    value: '--enable-cumem-allocator',
     options: []
   },
-
   {
     label: '--model-impl',
     value: '--model-impl',
-    options: ['auto', 'vllm', 'transformers']
+    options: ['auto', 'terratorch', 'transformers', 'vllm']
   },
+  {
+    label: '--override-attention-dtype',
+    value: '--override-attention-dtype',
+    options: []
+  },
+  {
+    label: '--logits-processors',
+    value: '--logits-processors',
+    options: []
+  },
+  {
+    label: '--io-processor-plugin',
+    value: '--io-processor-plugin',
+    options: []
+  },
+  {
+    label: '--renderer-num-workers',
+    value: '--renderer-num-workers',
+    options: []
+  },
+
+  // LoadConfig
+  {
+    label: '--load-format',
+    value: '--load-format',
+    options: [
+      'auto',
+      'pt',
+      'safetensors',
+      'instanttensor',
+      'npcache',
+      'dummy',
+      'tensorizer',
+      'runai_streamer',
+      'runai_streamer_sharded',
+      'bitsandbytes',
+      'sharded_state',
+      'mistral',
+      'modelexpress'
+    ]
+  },
+  {
+    label: '--download-dir',
+    value: '--download-dir',
+    options: []
+  },
+  {
+    label: '--safetensors-load-strategy',
+    value: '--safetensors-load-strategy',
+    options: ['eager', 'lazy', 'prefetch', 'torchao', 'None']
+  },
+  {
+    label: '--safetensors-prefetch-num-threads',
+    value: '--safetensors-prefetch-num-threads',
+    options: []
+  },
+  {
+    label: '--safetensors-prefetch-block-size',
+    value: '--safetensors-prefetch-block-size',
+    options: []
+  },
+  {
+    label: '--model-loader-extra-config',
+    value: '--model-loader-extra-config',
+    options: []
+  },
+  {
+    label: '--ignore-patterns',
+    value: '--ignore-patterns',
+    options: []
+  },
+  {
+    label: '--use-tqdm-on-load',
+    value: '--use-tqdm-on-load',
+    options: []
+  },
+  {
+    label: '--pt-load-map-location',
+    value: '--pt-load-map-location',
+    options: []
+  },
+
+  // AttentionConfig
+  {
+    label: '--attention-backend',
+    value: '--attention-backend',
+    options: []
+  },
+
+  // MambaConfig
+  {
+    label: '--mamba-backend',
+    value: '--mamba-backend',
+    options: []
+  },
+  {
+    label: '--enable-mamba-cache-stochastic-rounding',
+    value: '--enable-mamba-cache-stochastic-rounding',
+    options: []
+  },
+  {
+    label: '--mamba-cache-philox-rounds',
+    value: '--mamba-cache-philox-rounds',
+    options: []
+  },
+
+  // StructuredOutputsConfig
+  {
+    label: '--reasoning-parser',
+    value: '--reasoning-parser',
+    options: []
+  },
+  {
+    label: '--reasoning-parser-plugin',
+    value: '--reasoning-parser-plugin',
+    options: []
+  },
+
+  // ParallelConfig
   {
     label: '--distributed-executor-backend',
     value: '--distributed-executor-backend',
-    options: ['ray', 'mp', 'uni', 'external_launcher']
+    options: ['external_launcher', 'mp', 'ray', 'uni']
   },
   {
-    label: '--worker-use-ray',
-    value: '--worker-use-ray',
+    label: '--pipeline-parallel-size',
+    value: '--pipeline-parallel-size',
+    options: []
+  },
+  {
+    label: '--master-addr',
+    value: '--master-addr',
+    options: []
+  },
+  {
+    label: '--master-port',
+    value: '--master-port',
+    options: []
+  },
+  {
+    label: '--nnodes',
+    value: '--nnodes',
+    options: []
+  },
+  {
+    label: '--node-rank',
+    value: '--node-rank',
+    options: []
+  },
+  {
+    label: '--distributed-timeout-seconds',
+    value: '--distributed-timeout-seconds',
+    options: []
+  },
+  {
+    label: '--cpu-distributed-timeout-seconds',
+    value: '--cpu-distributed-timeout-seconds',
+    options: []
+  },
+  {
+    label: '--numa-bind',
+    value: '--numa-bind',
+    options: []
+  },
+  {
+    label: '--numa-bind-nodes',
+    value: '--numa-bind-nodes',
+    options: []
+  },
+  {
+    label: '--numa-bind-cpus',
+    value: '--numa-bind-cpus',
+    options: []
+  },
+  {
+    label: '--tensor-parallel-size',
+    value: '--tensor-parallel-size',
+    options: []
+  },
+  {
+    label: '--decode-context-parallel-size',
+    value: '--decode-context-parallel-size',
+    options: []
+  },
+  {
+    label: '--dcp-comm-backend',
+    value: '--dcp-comm-backend',
+    options: ['a2a', 'ag_rs']
+  },
+  {
+    label: '--dcp-kv-cache-interleave-size',
+    value: '--dcp-kv-cache-interleave-size',
+    options: []
+  },
+  {
+    label: '--cp-kv-cache-interleave-size',
+    value: '--cp-kv-cache-interleave-size',
+    options: []
+  },
+  {
+    label: '--prefill-context-parallel-size',
+    value: '--prefill-context-parallel-size',
     options: []
   },
   {
@@ -328,21 +725,124 @@ const options: BackendParameter[] = [
     value: '--data-parallel-size',
     options: []
   },
+  {
+    label: '--data-parallel-rank',
+    value: '--data-parallel-rank',
+    options: []
+  },
+  {
+    label: '--data-parallel-start-rank',
+    value: '--data-parallel-start-rank',
+    options: []
+  },
+  {
+    label: '--data-parallel-size-local',
+    value: '--data-parallel-size-local',
+    options: []
+  },
+  {
+    label: '--data-parallel-address',
+    value: '--data-parallel-address',
+    options: []
+  },
+  {
+    label: '--data-parallel-rpc-port',
+    value: '--data-parallel-rpc-port',
+    options: []
+  },
+  {
+    label: '--data-parallel-backend',
+    value: '--data-parallel-backend',
+    options: ['mp', 'ray']
+  },
+  {
+    label: '--data-parallel-hybrid-lb',
+    value: '--data-parallel-hybrid-lb',
+    options: []
+  },
+  {
+    label: '--data-parallel-external-lb',
+    value: '--data-parallel-external-lb',
+    options: []
+  },
+  {
+    label: '--data-parallel-multi-port-external-lb',
+    value: '--data-parallel-multi-port-external-lb',
+    options: []
+  },
   {
     label: '--enable-expert-parallel',
     value: '--enable-expert-parallel',
     options: []
   },
   {
-    label: '--pipeline-parallel-size',
-    value: '--pipeline-parallel-size',
+    label: '--enable-ep-weight-filter',
+    value: '--enable-ep-weight-filter',
     options: []
   },
   {
-    label: '--tensor-parallel-size',
-    value: '--tensor-parallel-size',
+    label: '--all2all-backend',
+    value: '--all2all-backend',
+    options: [
+      'allgather_reducescatter',
+      'deepep_high_throughput',
+      'deepep_low_latency',
+      'deepep_v2',
+      'flashinfer_all2allv',
+      'flashinfer_nvlink_one_sided',
+      'flashinfer_nvlink_two_sided',
+      'mori_high_throughput',
+      'mori_low_latency',
+      'naive',
+      'nixl_ep',
+      'pplx'
+    ]
+  },
+  {
+    label: '--enable-dbo',
+    value: '--enable-dbo',
     options: []
   },
+  {
+    label: '--ubatch-size',
+    value: '--ubatch-size',
+    options: []
+  },
+  {
+    label: '--enable-elastic-ep',
+    value: '--enable-elastic-ep',
+    options: []
+  },
+  {
+    label: '--dbo-decode-token-threshold',
+    value: '--dbo-decode-token-threshold',
+    options: []
+  },
+  {
+    label: '--dbo-prefill-token-threshold',
+    value: '--dbo-prefill-token-threshold',
+    options: []
+  },
+  {
+    label: '--disable-nccl-for-dp-synchronization',
+    value: '--disable-nccl-for-dp-synchronization',
+    options: []
+  },
+  {
+    label: '--enable-eplb',
+    value: '--enable-eplb',
+    options: []
+  },
+  {
+    label: '--eplb-config',
+    value: '--eplb-config',
+    options: []
+  },
+  {
+    label: '--expert-placement-strategy',
+    value: '--expert-placement-strategy',
+    options: ['linear', 'round_robin']
+  },
   {
     label: '--max-parallel-loading-workers',
     value: '--max-parallel-loading-workers',
@@ -353,10 +853,63 @@ const options: BackendParameter[] = [
     value: '--ray-workers-use-nsight',
     options: []
   },
+  {
+    label: '--disable-custom-all-reduce',
+    value: '--disable-custom-all-reduce',
+    options: []
+  },
+  {
+    label: '--worker-cls',
+    value: '--worker-cls',
+    options: []
+  },
+  {
+    label: '--worker-extension-cls',
+    value: '--worker-extension-cls',
+    options: []
+  },
+
+  // CacheConfig
   {
     label: '--block-size',
     value: '--block-size',
-    options: ['8', '16', '32', '64', '128']
+    options: ['1', '8', '16', '32', '64', '128']
+  },
+  {
+    label: '--gpu-memory-utilization',
+    value: '--gpu-memory-utilization',
+    options: []
+  },
+  {
+    label: '--kv-cache-memory-bytes',
+    value: '--kv-cache-memory-bytes',
+    options: []
+  },
+  {
+    label: '--kv-cache-dtype',
+    value: '--kv-cache-dtype',
+    options: [
+      'auto',
+      'bfloat16',
+      'float16',
+      'fp8',
+      'fp8_ds_mla',
+      'fp8_e4m3',
+      'fp8_e5m2',
+      'fp8_inc',
+      'fp8_per_token_head',
+      'int8_per_token_head',
+      'nvfp4',
+      'turboquant_3bit_nc',
+      'turboquant_4bit_nc',
+      'turboquant_k3v4_nc',
+      'turboquant_k8v4'
+    ]
+  },
+  {
+    label: '--num-gpu-blocks-override',
+    value: '--num-gpu-blocks-override',
+    options: []
   },
   {
     label: '--enable-prefix-caching',
@@ -366,182 +919,95 @@ const options: BackendParameter[] = [
   {
     label: '--prefix-caching-hash-algo',
     value: '--prefix-caching-hash-algo',
-    options: ['builtin', 'sha256']
+    options: ['sha256', 'sha256_cbor', 'xxhash', 'xxhash_cbor']
   },
   {
-    label: '--disable-sliding-window',
-    value: '--disable-sliding-window',
+    label: '--calculate-kv-scales',
+    value: '--calculate-kv-scales',
     options: []
   },
   {
-    label: '--use-v2-block-manager',
-    value: '--use-v2-block-manager',
+    label: '--kv-cache-dtype-skip-layers',
+    value: '--kv-cache-dtype-skip-layers',
     options: []
   },
   {
-    label: '--num-lookahead-slots',
-    value: '--num-lookahead-slots',
+    label: '--kv-sharing-fast-prefill',
+    value: '--kv-sharing-fast-prefill',
     options: []
   },
   {
-    label: '--seed',
-    value: '--seed',
+    label: '--mamba-cache-dtype',
+    value: '--mamba-cache-dtype',
+    options: ['auto', 'bfloat16', 'float16', 'float32']
+  },
+  {
+    label: '--mamba-ssm-cache-dtype',
+    value: '--mamba-ssm-cache-dtype',
+    options: ['auto', 'bfloat16', 'float16', 'float32']
+  },
+  {
+    label: '--mamba-block-size',
+    value: '--mamba-block-size',
     options: []
   },
   {
-    label: '--swap-space',
-    value: '--swap-space',
+    label: '--mamba-cache-mode',
+    value: '--mamba-cache-mode',
+    options: ['align', 'all', 'none']
+  },
+  {
+    label: '--kv-offloading-size',
+    value: '--kv-offloading-size',
     options: []
   },
+  {
+    label: '--kv-offloading-backend',
+    value: '--kv-offloading-backend',
+    options: ['lmcache', 'native']
+  },
+
+  // OffloadConfig
+  {
+    label: '--offload-backend',
+    value: '--offload-backend',
+    options: ['auto', 'prefetch', 'uva']
+  },
   {
     label: '--cpu-offload-gb',
     value: '--cpu-offload-gb',
     options: []
   },
   {
-    label: '--gpu-memory-utilization',
-    value: '--gpu-memory-utilization',
+    label: '--cpu-offload-params',
+    value: '--cpu-offload-params',
     options: []
   },
   {
-    label: '--num-gpu-blocks-override',
-    value: '--num-gpu-blocks-override',
+    label: '--offload-group-size',
+    value: '--offload-group-size',
     options: []
   },
   {
-    label: '--max-num-batched-tokens',
-    value: '--max-num-batched-tokens',
+    label: '--offload-num-in-group',
+    value: '--offload-num-in-group',
     options: []
   },
   {
-    label: '--max-num-partial-prefills',
-    value: '--max-num-partial-prefills',
+    label: '--offload-prefetch-step',
+    value: '--offload-prefetch-step',
+    options: []
+  },
+  {
+    label: '--offload-params',
+    value: '--offload-params',
     options: []
   },
 
+  // MultiModalConfig
   {
-    label: '--max-long-partial-prefills',
-    value: '--max-long-partial-prefills',
-    options: []
-  },
-
-  {
-    label: '--long-prefill-token-threshold',
-    value: '--long-prefill-token-threshold',
-    options: []
-  },
-  {
-    label: '--max-num-seqs',
-    value: '--max-num-seqs',
-    options: []
-  },
-  {
-    label: '--max-logprobs',
-    value: '--max-logprobs',
-    options: []
-  },
-  {
-    label: '--disable-log-stats',
-    value: '--disable-log-stats',
-    options: []
-  },
-  {
-    label: '--hf-overrides',
-    value: '--hf-overrides',
-    options: []
-  },
-  {
-    label: '--hf-token',
-    value: '--hf-token',
-    options: []
-  },
-
-  {
-    label: 'HF_TOKEN',
-    value: 'HF_TOKEN',
-    options: []
-  },
-  {
-    label: '--disable-mm-preprocessor-cache',
-    value: '--disable-mm-preprocessor-cache',
-    options: []
-  },
-  {
-    label: '--quantization',
-    value: '--quantization',
-    options: [
-      'aqlm',
-      'awq',
-      'deepspeedfp',
-      'tpu_int8',
-      'fp8',
-      'ptpc_fp8',
-      'fbgemm_fp8',
-      'modelopt',
-      'nvfp4',
-      'marlin',
-      'gguf',
-      'gptq_marlin_24',
-      'gptq_marlin',
-      'awq_marlin',
-      'gptq',
-      'compressed-tensors',
-      'bitsandbytes',
-      'qqq',
-      'hqq',
-      'experts_int8',
-      'neuron_quant',
-      'ipex',
-      'quark',
-      'moe_wna16',
-      'torchao',
-      'ascend',
-      'None'
-    ]
-  },
-  {
-    label: '--enable-lora-bias',
-    value: '--enable-lora-bias',
-    options: []
-  },
-  {
-    label: '--rope-theta',
-    value: '--rope-theta',
-    options: []
-  },
-  {
-    label: '--enforce-eager',
-    value: '--enforce-eager',
-    options: []
-  },
-  {
-    label: '--max-context-len-to-capture',
-    value: '--max-context-len-to-capture',
-    options: []
-  },
-  {
-    label: '--max-seq-len-to-capture',
-    value: '--max-seq-len-to-capture',
-    options: []
-  },
-  {
-    label: '--disable-custom-all-reduce',
-    value: '--disable-custom-all-reduce',
-    options: []
-  },
-  {
-    label: '--tokenizer-pool-size',
-    value: '--tokenizer-pool-size',
-    options: []
-  },
-  {
-    label: '--tokenizer-pool-type',
-    value: '--tokenizer-pool-type',
-    options: []
-  },
-  {
-    label: '--tokenizer-pool-extra-config',
-    value: '--tokenizer-pool-extra-config',
+    label: '--language-model-only',
+    value: '--language-model-only',
     options: []
   },
   {
@@ -549,11 +1015,93 @@ const options: BackendParameter[] = [
     value: '--limit-mm-per-prompt',
     options: []
   },
+  {
+    label: '--enable-mm-embeds',
+    value: '--enable-mm-embeds',
+    options: []
+  },
+  {
+    label: '--media-io-kwargs',
+    value: '--media-io-kwargs',
+    options: []
+  },
   {
     label: '--mm-processor-kwargs',
     value: '--mm-processor-kwargs',
     options: []
   },
+  {
+    label: '--mm-processor-cache-gb',
+    value: '--mm-processor-cache-gb',
+    options: []
+  },
+  {
+    label: '--mm-processor-cache-type',
+    value: '--mm-processor-cache-type',
+    options: ['lru', 'shm']
+  },
+  {
+    label: '--mm-shm-cache-max-object-size-mb',
+    value: '--mm-shm-cache-max-object-size-mb',
+    options: []
+  },
+  {
+    label: '--mm-encoder-only',
+    value: '--mm-encoder-only',
+    options: []
+  },
+  {
+    label: '--mm-encoder-tp-mode',
+    value: '--mm-encoder-tp-mode',
+    options: ['data', 'weights']
+  },
+  {
+    label: '--mm-encoder-attn-backend',
+    value: '--mm-encoder-attn-backend',
+    options: []
+  },
+  {
+    label: '--mm-encoder-attn-dtype',
+    value: '--mm-encoder-attn-dtype',
+    options: ['fp8', 'None']
+  },
+  {
+    label: '--mm-encoder-fp8-scale-path',
+    value: '--mm-encoder-fp8-scale-path',
+    options: []
+  },
+  {
+    label: '--mm-encoder-fp8-scale-save-path',
+    value: '--mm-encoder-fp8-scale-save-path',
+    options: []
+  },
+  {
+    label: '--mm-encoder-fp8-scale-save-margin',
+    value: '--mm-encoder-fp8-scale-save-margin',
+    options: []
+  },
+  {
+    label: '--interleave-mm-strings',
+    value: '--interleave-mm-strings',
+    options: []
+  },
+  {
+    label: '--skip-mm-profiling',
+    value: '--skip-mm-profiling',
+    options: []
+  },
+  {
+    label: '--video-pruning-rate',
+    value: '--video-pruning-rate',
+    options: []
+  },
+  {
+    label: '--mm-tensor-ipc',
+    value: '--mm-tensor-ipc',
+    options: ['direct_rpc', 'torch_shm']
+  },
+
+  // LoRAConfig
   {
     label: '--enable-lora',
     value: '--enable-lora',
@@ -567,21 +1115,16 @@ const options: BackendParameter[] = [
   {
     label: '--max-lora-rank',
     value: '--max-lora-rank',
-    options: []
-  },
-  {
-    label: '--lora-extra-vocab-size',
-    value: '--lora-extra-vocab-size',
-    options: []
+    options: ['1', '8', '16', '32', '64', '128', '256', '320', '512']
   },
   {
     label: '--lora-dtype',
     value: '--lora-dtype',
-    options: ['auto', 'float16', 'bfloat16']
+    options: []
   },
   {
-    label: '--long-lora-scaling-factors',
-    value: '--long-lora-scaling-factors',
+    label: '--enable-tower-connector-lora',
+    value: '--enable-tower-connector-lora',
     options: []
   },
   {
@@ -595,90 +1138,32 @@ const options: BackendParameter[] = [
     options: []
   },
   {
-    label: '--enable-prompt-adapter',
-    value: '--enable-prompt-adapter',
+    label: '--lora-target-modules',
+    value: '--lora-target-modules',
     options: []
   },
   {
-    label: '--max-prompt-adapters',
-    value: '--max-prompt-adapters',
+    label: '--default-mm-loras',
+    value: '--default-mm-loras',
     options: []
   },
   {
-    label: '--max-prompt-adapter-token',
-    value: '--max-prompt-adapter-token',
+    label: '--specialize-active-lora',
+    value: '--specialize-active-lora',
     options: []
   },
   {
-    label: '--device',
-    value: '--device',
-    options: ['auto', 'cuda', 'neuron', 'cpu', 'tpu', 'xpu', 'hpu']
-  },
-  {
-    label: '--speculative-config',
-    value: '--speculative-config',
-    options: []
-  },
-  {
-    label: '--num-scheduler-steps',
-    value: '--num-scheduler-steps',
-    options: []
-  },
-  {
-    label: '--multi-step-stream-outputs',
-    value: '--multi-step-stream-outputs',
-    options: []
-  },
-  {
-    label: '--scheduler-delay-factor',
-    value: '--scheduler-delay-factor',
-    options: []
-  },
-  {
-    label: '--enable-chunked-prefill',
-    value: '--enable-chunked-prefill',
-    options: []
-  },
-  {
-    label: '--model-loader-extra-config',
-    value: '--model-loader-extra-config',
-    options: []
-  },
-  {
-    label: '--scheduler-cls',
-    value: '--scheduler-cls',
-    options: []
-  },
-  {
-    label: '--use-tqdm-on-load',
-    value: '--use-tqdm-on-load',
-    options: []
-  },
-  {
-    label: '--ignore-patterns',
-    value: '--ignore-patterns',
+    label: '--enable-mixed-moe-lora-format',
+    value: '--enable-mixed-moe-lora-format',
     options: []
   },
+
+  // ObservabilityConfig
   {
     label: '--show-hidden-metrics-for-version',
     value: '--show-hidden-metrics-for-version',
     options: []
   },
-  {
-    label: '--preemption-mode',
-    value: '--preemption-mode',
-    options: []
-  },
-  {
-    label: '--served-model-name',
-    value: '--served-model-name',
-    options: []
-  },
-  {
-    label: '--qlora-adapter-name-or-path',
-    value: '--qlora-adapter-name-or-path',
-    options: []
-  },
   {
     label: '--otlp-traces-endpoint',
     value: '--otlp-traces-endpoint',
@@ -687,57 +1172,68 @@ const options: BackendParameter[] = [
   {
     label: '--collect-detailed-traces',
     value: '--collect-detailed-traces',
+    options: ['all', 'model', 'worker', 'None']
+  },
+  {
+    label: '--kv-cache-metrics',
+    value: '--kv-cache-metrics',
     options: []
   },
   {
-    label: '--disable-async-output-proc',
-    value: '--disable-async-output-proc',
+    label: '--kv-cache-metrics-sample',
+    value: '--kv-cache-metrics-sample',
     options: []
   },
   {
-    label: '--override-neuron-config',
-    value: '--override-neuron-config',
+    label: '--cudagraph-metrics',
+    value: '--cudagraph-metrics',
     options: []
   },
   {
-    label: '--compilation-config',
-    value: '--compilation-config',
+    label: '--enable-layerwise-nvtx-tracing',
+    value: '--enable-layerwise-nvtx-tracing',
     options: []
   },
   {
-    label: '--override-pooler-config',
-    value: '--override-pooler-config',
+    label: '--enable-mfu-metrics',
+    value: '--enable-mfu-metrics',
     options: []
   },
   {
-    label: '--kv-transfer-config',
-    value: '--kv-transfer-config',
+    label: '--enable-logging-iteration-details',
+    value: '--enable-logging-iteration-details',
     options: []
   },
   {
-    label: '--worker-cls',
-    value: '--worker-cls',
-    options: []
-  },
-  {
-    label: '--override-generation-config',
-    value: '--override-generation-config',
-    options: []
-  },
-  {
-    label: '--enable-sleep-mode',
-    value: '--enable-sleep-mode',
+    label: '--jit-monitor-verbose',
+    value: '--jit-monitor-verbose',
     options: []
   },
 
+  // SchedulerConfig
   {
-    label: '--calculate-kv-scales',
-    value: '--calculate-kv-scales',
+    label: '--max-num-batched-tokens',
+    value: '--max-num-batched-tokens',
     options: []
   },
   {
-    label: '--generation-config',
-    value: '--generation-config',
+    label: '--max-num-seqs',
+    value: '--max-num-seqs',
+    options: []
+  },
+  {
+    label: '--max-num-partial-prefills',
+    value: '--max-num-partial-prefills',
+    options: []
+  },
+  {
+    label: '--max-long-partial-prefills',
+    value: '--max-long-partial-prefills',
+    options: []
+  },
+  {
+    label: '--long-prefill-token-threshold',
+    value: '--long-prefill-token-threshold',
     options: []
   },
   {
@@ -746,33 +1242,199 @@ const options: BackendParameter[] = [
     options: ['fcfs', 'priority']
   },
   {
-    label: '--disable-log-requests',
-    value: '--disable-log-requests',
+    label: '--enable-chunked-prefill',
+    value: '--enable-chunked-prefill',
     options: []
   },
   {
-    label: '--additional-config',
-    value: '--additional-config',
+    label: '--disable-chunked-mm-input',
+    value: '--disable-chunked-mm-input',
     options: []
   },
   {
-    label: '--max-log-len',
-    value: '--max-log-len',
+    label: '--scheduler-cls',
+    value: '--scheduler-cls',
     options: []
   },
   {
-    label: '--disable-fastapi-docs',
-    value: '--disable-fastapi-docs',
+    label: '--scheduler-reserve-full-isl',
+    value: '--scheduler-reserve-full-isl',
     options: []
   },
   {
-    label: '--enable-prompt-tokens-details',
-    value: '--enable-prompt-tokens-details',
+    label: '--watermark',
+    value: '--watermark',
     options: []
   },
   {
-    label: '--enable-server-load-tracking',
-    value: '--enable-server-load-tracking',
+    label: '--prefill-schedule-interval',
+    value: '--prefill-schedule-interval',
+    options: []
+  },
+  {
+    label: '--disable-hybrid-kv-cache-manager',
+    value: '--disable-hybrid-kv-cache-manager',
+    options: []
+  },
+  {
+    label: '--async-scheduling',
+    value: '--async-scheduling',
+    options: []
+  },
+  {
+    label: '--stream-interval',
+    value: '--stream-interval',
+    options: []
+  },
+
+  // CompilationConfig
+  {
+    label: '--cudagraph-capture-sizes',
+    value: '--cudagraph-capture-sizes',
+    options: []
+  },
+  {
+    label: '--max-cudagraph-capture-size',
+    value: '--max-cudagraph-capture-size',
+    options: []
+  },
+
+  // KernelConfig
+  {
+    label: '--ir-op-priority',
+    value: '--ir-op-priority',
+    options: []
+  },
+  {
+    label: '--enable-flashinfer-autotune',
+    value: '--enable-flashinfer-autotune',
+    options: []
+  },
+  {
+    label: '--moe-backend',
+    value: '--moe-backend',
+    options: [
+      'aiter',
+      'auto',
+      'cutlass',
+      'deep_gemm',
+      'deep_gemm_mega_moe',
+      'emulation',
+      'flashinfer_b12x',
+      'flashinfer_cutedsl',
+      'flashinfer_cutlass',
+      'flashinfer_trtllm',
+      'flydsl',
+      'humming',
+      'marlin',
+      'triton',
+      'triton_unfused'
+    ]
+  },
+  {
+    label: '--linear-backend',
+    value: '--linear-backend',
+    options: [
+      'aiter',
+      'auto',
+      'conch',
+      'cutlass',
+      'deep_gemm',
+      'emulation',
+      'exllama',
+      'fbgemm',
+      'flashinfer_b12x',
+      'flashinfer_cudnn',
+      'flashinfer_cutlass',
+      'flashinfer_trtllm',
+      'machete',
+      'marlin',
+      'torch',
+      'triton'
+    ]
+  },
+
+  // VllmConfig
+  {
+    label: '--speculative-config',
+    value: '--speculative-config',
+    options: []
+  },
+  {
+    label: '--spec-method',
+    value: '--spec-method',
+    options: [
+      'custom_class',
+      'deepseek_mtp',
+      'dflash',
+      'draft_model',
+      'eagle',
+      'eagle3',
+      'ernie_mtp',
+      'exaone4_5_mtp',
+      'exaone_moe_mtp',
+      'extract_hidden_states',
+      'gemma4_mtp',
+      'glm4_moe_lite_mtp',
+      'glm4_moe_mtp',
+      'glm_ocr_mtp',
+      'hy_v3_mtp',
+      'longcat_flash_mtp',
+      'medusa',
+      'mimo_mtp',
+      'mimo_v2_mtp',
+      'minimax_m3_mtp',
+      'mlp_speculator',
+      'mtp',
+      'nemotron_h_mtp',
+      'ngram',
+      'ngram_gpu',
+      'pangu_ultra_moe_mtp',
+      'qwen3_5_mtp',
+      'qwen3_next_mtp',
+      'step3p5_mtp',
+      'suffix',
+      'None'
+    ]
+  },
+  {
+    label: '--spec-model',
+    value: '--spec-model',
+    options: []
+  },
+  {
+    label: '--spec-tokens',
+    value: '--spec-tokens',
+    options: []
+  },
+  {
+    label: '--diffusion-config',
+    value: '--diffusion-config',
+    options: []
+  },
+  {
+    label: '--kv-transfer-config',
+    value: '--kv-transfer-config',
+    options: []
+  },
+  {
+    label: '--kv-events-config',
+    value: '--kv-events-config',
+    options: []
+  },
+  {
+    label: '--ec-transfer-config',
+    value: '--ec-transfer-config',
+    options: []
+  },
+  {
+    label: '--compilation-config',
+    value: '--compilation-config',
+    options: []
+  },
+  {
+    label: '--attention-config',
+    value: '--attention-config',
     options: []
   }
 ];