diff --git a/.modelhub_state/recovery_intents.jsonl b/.modelhub_state/recovery_intents.jsonl
index d1045636f3..b11731fbc9 100644
--- a/.modelhub_state/recovery_intents.jsonl
+++ b/.modelhub_state/recovery_intents.jsonl
@@ -3477,7 +3477,7 @@
 {"batchId": "24940d0f108246c3baf71dcf0464bf5e", "completedAt": "2026-10-13T19:03:26.107059+00:00", "configFingerprint": "16f74ea0f409861ab0c8169a25d9484cb482e27d6a3ad7c529a70b7cdbfabee3", "configSource": "modelhub_live", "createdAt": "2026-10-13T19:01:58.232433+00:00", "framework": "vllm_fix_tokenizer", "intentId": "124e6e7d60ff40e4a24cfa80a1170c2c", "lastModified": "2026-10-13T18:27:56+00:00", "modelAddress": "https://modelscope.cn/models/LiquidAI/LFM2.5-24B-A2B-Base", "reason": null, "reconciledAt": "2026-10-13T19:03:30.242896+00:00", "repoId": "LiquidAI/LFM2.5-24B-A2B-Base", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "recovered_active", "targetGpu": "Kunlunxin_p-800", "taskId": "5517740", "taskType": "text-generation"}
 {"batchId": "24940d0f108246c3baf71dcf0464bf5e", "completedAt": "2026-10-13T19:03:26.107071+00:00", "configFingerprint": "16f74ea0f409861ab0c8169a25d9484cb482e27d6a3ad7c529a70b7cdbfabee3", "configSource": "modelhub_live", "createdAt": "2026-10-13T19:01:58.232530+00:00", "framework": "vllm_fix_tokenizer", "intentId": "1a94a77a92c6451ba17d23a4f60b2fad", "lastModified": "2026-10-13T18:39:46+00:00", "modelAddress": "https://modelscope.cn/models/LiquidAI/LFM2.5-24B-A2B-MLX-8bit", "reason": null, "reconciledAt": "2026-10-13T19:03:30.243090+00:00", "repoId": "LiquidAI/LFM2.5-24B-A2B-MLX-8bit", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "recovered_active", "targetGpu": "Kunlunxin_p-800", "taskId": "5517741", "taskType": "text-generation"}
 {"batchId": "24940d0f108246c3baf71dcf0464bf5e", "completedAt": "2026-10-13T19:03:26.107074+00:00", "configFingerprint": "16f74ea0f409861ab0c8169a25d9484cb482e27d6a3ad7c529a70b7cdbfabee3", "configSource": "modelhub_live", "createdAt": "2026-10-13T19:01:58.232577+00:00", "framework": "vllm_fix_tokenizer", "intentId": "a7b965219c474cfe9e4238ce72c296f6", "lastModified": "2026-10-13T18:29:48+00:00", "modelAddress": "https://modelscope.cn/models/LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "reason": null, "reconciledAt": "2026-10-13T19:03:30.242460+00:00", "repoId": "LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "recovered_active", "targetGpu": "Kunlunxin_p-800", "taskId": "5517739", "taskType": "text-generation"}
-{"batchId": "f3ca411acc254af9b0c6073547ccc96b", "configFingerprint": "36ec94e289a8da5fd70b98a2f401fc5159740daf9e2a076da4fd490f34bcbc0b", "configSource": "modelhub_live", "createdAt": "2026-10-13T19:06:03.572957+00:00", "framework": "vllm-mlu", "intentId": "174f409c19364b2d94f14adf7df8efc0", "lastModified": "2026-10-13T18:29:48+00:00", "modelAddress": "https://modelscope.cn/models/LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "repoId": "LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "safeConfigVector": {"dtype": "-", "gpuNum": 1}, "status": "pending", "targetGpu": "Cambricon_mlu-370-x4", "taskType": "text-generation"}
+{"batchId": "f3ca411acc254af9b0c6073547ccc96b", "completedAt": "2026-10-13T19:07:11.951904+00:00", "configFingerprint": "36ec94e289a8da5fd70b98a2f401fc5159740daf9e2a076da4fd490f34bcbc0b", "configSource": "modelhub_live", "createdAt": "2026-10-13T19:06:03.572957+00:00", "framework": "vllm-mlu", "intentId": "174f409c19364b2d94f14adf7df8efc0", "lastModified": "2026-10-13T18:29:48+00:00", "modelAddress": "https://modelscope.cn/models/LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "reason": null, "repoId": "LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "safeConfigVector": {"dtype": "-", "gpuNum": 1}, "status": "submitted", "targetGpu": "Cambricon_mlu-370-x4", "taskId": "5517780", "taskType": "text-generation"}
 {"batchId": "64ef2259965a4a47875e42e124456e7a", "completedAt": "2026-10-10T08:22:29.829255+00:00", "configFingerprint": "014e4c7848d2739d8b20b385d6b85827a2fc5723d3b5a17c03afbd67c56275d9", "configSource": "modelhub_live", "createdAt": "2026-10-10T08:19:42.281118+00:00", "framework": "llamacpp", "intentId": "1c4c57e0fcb04deda8bccda4d928eff0", "repoId": "GnLOLot/MiniCPM5-2B-Claude-Fable5-1-Thinking-Agentic-GGUF", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "uniqueness_rejected", "targetGpu": "Iluvatar_bi-150", "taskId": null, "taskType": "text-generation"}
 {"batchId": "04247231908a4293bb90fcfb6ceb1b21", "completedAt": "2026-10-10T03:37:10.311161+00:00", "configFingerprint": "41b32e99f4cd3b280f48beb66666c48e543570dbfdee61682247438440dae4f3", "configSource": "modelhub_live", "createdAt": "2026-10-10T03:35:16.235834+00:00", "framework": "llamacpp", "intentId": "7ec4fc2708014d83955beed36401eb73", "repoId": "GnLOLot/MiniCPM5-2B-Claude-Fable5-1-Thinking-Agentic-GGUF", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 2048}, "status": "uniqueness_rejected", "targetGpu": "Mthreads_s4000", "taskId": null, "taskType": "text-generation"}
 {"batchId": "defc40efc99547dea797f26d6a4c1a65", "completedAt": "2026-10-10T03:33:51.337890+00:00", "configFingerprint": "b6b49b497726342799df1608aa88399dc2d2712f97d7f1eb9fbc89c5b78f2c4c", "configSource": "modelhub_live", "createdAt": "2026-10-10T03:32:02.289053+00:00", "framework": "llamacpp", "intentId": "4dc3d20711774633bcaa193b5c6032ab", "repoId": "GnLOLot/MiniCPM5-2B-Claude-Fable5-1-Thinking-Agentic-GGUF", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "uniqueness_rejected", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
diff --git a/.modelhub_state/routing_intelligence.json b/.modelhub_state/routing_intelligence.json
index ca4a3e2114..4dfcc6e538 100644
--- a/.modelhub_state/routing_intelligence.json
+++ b/.modelhub_state/routing_intelligence.json
@@ -1,6 +1,6 @@
 {
   "acceptedByCategory": {
-    "unified_success_first": 3616
+    "unified_success_first": 3617
   },
   "acceptedByRoute": {
     "text-generation|Ascend_910-b3|llamacpp": 31,
@@ -13,7 +13,7 @@
     "text-generation|Biren_166m|vllm_fix_tokenizer": 252,
     "text-generation|Cambricon_mlu-370-x4|vllm": 73,
     "text-generation|Cambricon_mlu-370-x4|vllm-customized": 15,
-    "text-generation|Cambricon_mlu-370-x4|vllm-mlu": 180,
+    "text-generation|Cambricon_mlu-370-x4|vllm-mlu": 181,
     "text-generation|Cambricon_mlu-370-x8|vllm": 58,
     "text-generation|Cambricon_mlu-370-x8|vllm-customized": 71,
     "text-generation|Cambricon_mlu-370-x8|vllm-mlu": 187,
@@ -39,8 +39,8 @@
     "text-generation|hygon_k100-ai|vllm": 35,
     "text-generation|hygon_k100-ai|vllm-patch-tokenizer": 232
   },
-  "acceptedSinceRefresh": 3616,
-  "acceptedTotal": 3616,
-  "generatedAt": "2026-10-13T19:03:26.045814+00:00",
+  "acceptedSinceRefresh": 3617,
+  "acceptedTotal": 3617,
+  "generatedAt": "2026-10-13T19:07:11.891721+00:00",
   "version": 1
 }
diff --git a/ledger/submissions.jsonl b/ledger/submissions.jsonl
index b324b4b499..b65a9e1de7 100644
--- a/ledger/submissions.jsonl
+++ b/ledger/submissions.jsonl
@@ -1111,3 +1111,4 @@
 {"framework": "vllm_tokenizer_patch", "modelAddress": "https://modelscope.cn/models/XingChen-AGI/Xing4.0-29B-A4B", "modelId": "XingChen-AGI/Xing4.0-29B-A4B", "submitTime": "2026-10-13T12:30:24.554632+00:00", "targetGpu": "MetaX_c-500", "taskId": "5512351", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-vllm-tokenizer-patch-metax-c-500"}
 {"framework": "vllm_tokenizer_patch", "modelAddress": "https://modelscope.cn/models/XingChen-AGI/Xing4.0-29B-A4B", "modelId": "XingChen-AGI/Xing4.0-29B-A4B", "submitTime": "2026-10-13T12:52:58.552750+00:00", "targetGpu": "Ascend_910-b3", "taskId": "5512698", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-vllm-tokenizer-patch-ascend-910-b3"}
 {"framework": "vllm-patch-tokenizer", "modelAddress": "https://modelscope.cn/models/XingChen-AGI/Xing4.0-29B-A4B", "modelId": "XingChen-AGI/Xing4.0-29B-A4B", "submitTime": "2026-10-13T13:12:18.147765+00:00", "targetGpu": "hygon_k100-ai", "taskId": "5513114", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-vllm-patch-tokenizer-hygon-k100-ai"}
+{"framework": "vllm-mlu", "modelAddress": "https://modelscope.cn/models/LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "modelId": "LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "submitTime": "2026-10-13T19:06:05.151931+00:00", "targetGpu": "Cambricon_mlu-370-x4", "taskId": "5517780", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-vllm-mlu-cambricon-mlu-370-x4"}
diff --git a/manifest.json b/manifest.json
index 492158813b..937c482097 100644
--- a/manifest.json
+++ b/manifest.json
@@ -10,16 +10,16 @@
     ".modelhub_state/queue_cleanup_latest.json": "a853c3ecd05b6eb13dc78028071c981c78c9adf73c9831a523b9562a8bdb1610",
     ".modelhub_state/recent_outcomes.jsonl": "d5f3321c7c835be11fc790c6b04936eee563f577da3818cfa5dc8aa7c8332239",
     ".modelhub_state/recovery_active_tasks.jsonl": "07f2c3f1104600beab433d6eb5e80eff6fcd72f80b1b0d39a10487f630e5dea8",
-    ".modelhub_state/recovery_intents.jsonl": "351e81684dd926201677149f6a767c720a265b194c387dc0a83ea4d7ce818ecc",
-    ".modelhub_state/routing_intelligence.json": "ff2bfa6cd5d965eace6f902b0d2462e5f242c06e45090b593417857865b76686",
+    ".modelhub_state/recovery_intents.jsonl": "7ecc1d79be8cb374b59905f906202a6148d6a3dda94dbd2931bde938305c9983",
+    ".modelhub_state/routing_intelligence.json": "d6b6815d22e562f0f87641aec276d81999c3d420cbcb4e25693eef18b0250c9d",
     ".modelhub_state/submission_exclusions.jsonl": "660ad77b924d3b0cc76f46aacd1f91a924d13a02bc68057eb5a964647441283b",
     ".modelhub_state/worker_crashes.jsonl": "a6b56caed5c7afeb803f8701393cf4660f7e9799eb5ff84af826a142e16d9b00",
-    "ledger/submissions.jsonl": "1c7fd981e8d905739f530dbb5fee41cb777960fbfdfa8b7f9920787c465abe12",
-    "outcomes/submissions.jsonl": "38a6431041f08743c9a4e819a74b1e0784bfbfff1d8df59cb97f7d6a79bba669"
+    "ledger/submissions.jsonl": "be6bed3b9029843b5eed91dc01d74bcd4cc20369bae88b4d7a2f2835c54e7251",
+    "outcomes/submissions.jsonl": "aa9836b3027182abcc5e20568eade22af25fd6764626bd6476cca80a0b013034"
   },
-  "generation": 24375,
-  "phase": "intent",
+  "generation": 24376,
+  "phase": "result",
   "schemaVersion": 1,
-  "updatedAt": "2026-10-13T19:06:03.735274+00:00",
+  "updatedAt": "2026-10-13T19:07:12.084643+00:00",
   "writerId": "0b4d13138a434e929040a457f5ff8e93"
 }
diff --git a/outcomes/submissions.jsonl b/outcomes/submissions.jsonl
index a623f165d5..f7bf8155af 100644
--- a/outcomes/submissions.jsonl
+++ b/outcomes/submissions.jsonl
@@ -962,3 +962,4 @@
 {"failReason": null, "framework": "vllm_fix_tokenizer", "lastSyncTime": null, "modelId": "LiquidAI/LFM2.5-24B-A2B-Base", "modelProfile": {"architectures": ["Lfm2MoeForCausalLM"], "configFingerprint": "16f74ea0f409861ab0c8169a25d9484cb482e27d6a3ad7c529a70b7cdbfabee3", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 47944143456, "estimatedRequiredGiB": 53.603, "gpuMemoryEvidence": {"memoryGiB": 100.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 100.0, "maximumRepositorySizeGiB": 83.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "lfm2_moe", "modelscopeFileSize": 47962770138, "modelscopeLicense": "other", "modelscopeParams": null, "modelscopeTags": ["license:other", "model_type:lfm2_moe", "library:transformer", "library:safetensors", "library:pytorch", "task:text-generation", "custom_tag:liquid", "custom_tag:lfm2.5", "custom_tag:edge", "custom_tag:moe"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 47962770138}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-13T19:02:00.548510+00:00", "targetGpu": "Kunlunxin_p-800", "taskId": "5517740", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm_fix_tokenizer", "lastSyncTime": null, "modelId": "LiquidAI/LFM2.5-24B-A2B-MLX-8bit", "modelProfile": {"architectures": ["Lfm2MoeForCausalLM"], "configFingerprint": "16f74ea0f409861ab0c8169a25d9484cb482e27d6a3ad7c529a70b7cdbfabee3", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 25484884997, "estimatedRequiredGiB": 28.502, "gpuMemoryEvidence": {"memoryGiB": 100.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 100.0, "maximumRepositorySizeGiB": 83.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "lfm2_moe", "modelscopeFileSize": null, "modelscopeLicense": "other", "modelscopeParams": null, "modelscopeTags": ["license:other", "model_type:lfm2_moe", "library:mlx", "library:safetensors", "library:pytorch", "task:text-generation", "custom_tag:liquid", "custom_tag:lfm2.5", "custom_tag:edge", "custom_tag:moe", "custom_tag:mlx"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 25502895279}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-13T19:02:00.146586+00:00", "targetGpu": "Kunlunxin_p-800", "taskId": "5517741", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm_fix_tokenizer", "lastSyncTime": null, "modelId": "LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "modelProfile": {"architectures": ["Lfm2MoeForCausalLM"], "configFingerprint": "16f74ea0f409861ab0c8169a25d9484cb482e27d6a3ad7c529a70b7cdbfabee3", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 19493321709, "estimatedRequiredGiB": 21.806, "gpuMemoryEvidence": {"memoryGiB": 100.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 100.0, "maximumRepositorySizeGiB": 83.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "lfm2_moe", "modelscopeFileSize": null, "modelscopeLicense": "other", "modelscopeParams": null, "modelscopeTags": ["license:other", "model_type:lfm2_moe", "library:mlx", "library:safetensors", "library:pytorch", "task:text-generation", "custom_tag:liquid", "custom_tag:lfm2.5", "custom_tag:edge", "custom_tag:moe", "custom_tag:mlx"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 19511331991}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-13T19:02:00.052611+00:00", "targetGpu": "Kunlunxin_p-800", "taskId": "5517739", "taskType": "text-generation", "verifyResult": null}
+{"failReason": null, "framework": "vllm-mlu", "lastSyncTime": null, "modelId": "LiquidAI/LFM2.5-24B-A2B-MLX-6bit", "modelProfile": {"architectures": ["Lfm2MoeForCausalLM"], "configFingerprint": "36ec94e289a8da5fd70b98a2f401fc5159740daf9e2a076da4fd490f34bcbc0b", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 19493321709, "estimatedRequiredGiB": 21.806, "gpuMemoryEvidence": {"memoryGiB": 24.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 24.0, "maximumRepositorySizeGiB": 20.0, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "lfm2_moe", "modelscopeFileSize": null, "modelscopeLicense": "other", "modelscopeParams": null, "modelscopeTags": ["license:other", "model_type:lfm2_moe", "library:mlx", "library:safetensors", "library:pytorch", "task:text-generation", "custom_tag:liquid", "custom_tag:lfm2.5", "custom_tag:edge", "custom_tag:moe", "custom_tag:mlx"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 19511331991}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-13T19:06:05.151931+00:00", "targetGpu": "Cambricon_mlu-370-x4", "taskId": "5517780", "taskType": "text-generation", "verifyResult": null}