Recipe
qwen3-6-35b-a3b-nvfp4-rtx-pro-6000-blackwell-96gb-vllm-tp1
qwen3-6-35b-a3b-nvfp4-rtx-pro-6000-blackwell-96gb-vllm-tp1Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.
Record
- Status
- candidate
- Source
- localmaxxing
- Engine
- vllm
- Engine version
- 0.19.x-pr40898-cu130
- Accelerators
- 1
- Tensor parallel
- 1
- Context tokens
- 8,192
- Max concurrency
- 1
- chat
- unknown
- reasoning
- unknown
- tools
- unknown
- vision
- unknown
Hugging Face model card
Identity
https://huggingface.co/mmangkad/Qwen3.6-35B-A3B-NVFP4- Repository
- mmangkad/Qwen3.6-35B-A3B-NVFP4
- Status
- known
- Link type
- Exact Hub repository
Public Hugging Face repository confirmed by the Hub API.
Candidate: useful compatibility or speed evidence. The registry does not offer Run until promotion requirements are met.
Observed configuration
vllm
Evidence only · candidate · reference
Candidate evidence — not a Run contract
Observed source tokens
Mechanical split of the source command. Unverified against the engine CLI. The registry does not offer Run for this recipe.
vllmservemmangkad/Qwen3.6-35B-A3B-NVFP4--quantizationnvfp4--max-model-len8192--max-num-seqs1--gpu-memory-utilization0.35--kv-cache-dtypebfloat16--tool-call-parserqwen3_coder--enable-auto-tool-choice--speculative-config{method:dflash,num_speculative_tokens:4}--compilation-config{level:2,inductor_compile_config:{use_inductor_graph_partition:true}}
| Flag | Value |
|---|---|
--quantization | nvfp4 |
--max-model-len | 8192 |
--max-num-seqs | 1 |
--gpu-memory-utilization | 0.35 |
--kv-cache-dtype | bfloat16 |
--tool-call-parser | qwen3_coder |
--speculative-config | {method:dflash,num_speculative_tokens:4} |
--compilation-config | {level:2,inductor_compile_config:{use_inductor_graph_partition:true}} |
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status | Sweep |
|---|---|---|---|---|---|---|
| 1 | 8,192 | — | 506.2 | 113.4 | observed | qwen3-6-35b-a3b-nvfp4-rtx-pro-6000-blackwell-96gb-vllm-tp1-sweep |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- hardware count
- 1
- hardware id
- rtx-pro-6000-blackwell-96gb
- id
- qwen3-6-35b-a3b-nvfp4-rtx-pro-6000-blackwell-96gb-vllm-tp1
- model instance id
- mmangkad-qwen3-6-35b-a3b-nvfp4--nvfp4
- recipe source
- localmaxxing
- schema version
- local-ai-registry/v1
- speed sweep ids
- qwen3-6-35b-a3b-nvfp4-rtx-pro-6000-blackwell-96gb-vllm-tp1-sweep
- status
- candidate
capabilities
draft launch
- accelerator backend
- nvidia
- arguments
- --model, mmangkad/Qwen3.6-35B-A3B-NVFP4, --tensor-parallel-size, 1, --host, 0.0.0.0, --port, 8000, --max-model-len, 8192
- container port
- 8,000
- host port
- 8,000
- image
- vllm/vllm-openai@sha256:0a51ea5b4ae2dc5d81890e5173f54203d2a3ae0cfffe51b8fd2afd4391bfd967
- ipc
- host
- kind
- docker
- shm size
- 16g
- synthesized · generated at
- 2026-08-31T22:12:17Z
- synthesized · image provenance
- deepseek-fp8-rtx-pro-6000-blackwell-96gb-vllm-tp1
- synthesized · template
- vllm-openai-v1
mounts
| read only | target |
|---|---|
| No | /root/.cache/huggingface |
engine
- name
- vllm
- version
- 0.19.x-pr40898-cu130
serving
- max concurrency
- 1
- max context tokens
- 8,192
- tensor parallel
- 1
Provenance & metadata (3)
facts
- capabilities.chat · provenance · captured at
- 2026-08-30T09:10:02Z
capabilities.chat · reason capability-not-verified
capabilities.chat · state unknown
- capabilities.reasoning · provenance · captured at
- 2026-08-30T09:10:02Z
capabilities.reasoning · reason capability-not-verified
capabilities.reasoning · state unknown
- capabilities.tools · provenance · captured at
- 2026-08-30T09:10:02Z
capabilities.tools · reason capability-not-verified
capabilities.tools · state unknown
- capabilities.vision · provenance · captured at
- 2026-08-30T09:10:02Z
capabilities.vision · reason capability-not-verified
capabilities.vision · state unknown
- engine.graph mode · provenance · captured at
- 2026-08-30T09:10:02Z
engine.graph mode · reason runtime-detail-not-published
engine.graph mode · state unknown
- serving.kv cache tokens · provenance · captured at
- 2026-08-30T09:10:02Z
serving.kv cache tokens · reason kv-cache-capacity-not-published
serving.kv cache tokens · state unknown
metadata
- localmaxxing · hardware label
- RTX PRO 6000 Blackwell
- localmaxxing · notes
- Campaign champion — graph_partition + FAP compile (no max-autotune) on SM120 Blackwell. 2.14x over pre-compile baseline. BFCL 73.33% (22/30). DFlash speculative decoding N=4 + CUDA graphs + greedy (temp=0). 20/20 tool-call chain completed, 100% tool_call_valid_rate. Warmup=10 discarded runs before measurement.
- localmaxxing · observed command
- vllm serve mmangkad/Qwen3.6-35B-A3B-NVFP4 --quantization nvfp4 --max-model-len 8192 --max-num-seqs 1 --gpu-memory-utilization 0.35 --kv-cache-dtype bfloat16 --tool-call-parser qwen3_coder --enable-auto-tool-choice --speculative-config {"method":"dflash","num_speculative_tokens":4} --compilation-config {"level":2,"inductor_compile_config":{"use_inductor_graph_partition":true}}
- localmaxxing · run id
- cmopmgspw0002jv041yjbb6sq
- localmaxxing · tokenized · arguments
- vllm, serve, mmangkad/Qwen3.6-35B-A3B-NVFP4, --quantization, nvfp4, --max-model-len, 8192, --max-num-seqs, 1, --gpu-memory-utilization, 0.35, --kv-cache-dtype, bfloat16, --tool-call-parser, qwen3_coder, --enable-auto-tool-choice, --speculative-config, {method:dflash,num_speculative_tokens:4}, --compilation-config, {level:2,inductor_compile_config:{use_inductor_graph_partition:true}}
- localmaxxing · tokenized · fidelity
- faithful
provenance
- captured at
- 2026-08-30T09:10:02Z
sources
| captured at | kind | url |
|---|---|---|
| 2026-08-30T09:10:02Z | normalized-recipe | www.localmaxxing.com/en/runs/cmopmgspw0002jv041yjbb6sq ↗ |