Local AI Registry

Recipe

minimax-m2-7-int4-autoround-w4a16-intel-arc-pro-b70-32gb-vllm-tp4-uqa329e6

minimax-m2-7-int4-autoround-w4a16-intel-arc-pro-b70-32gb-vllm-tp4-uqa329e6
JSON APIFind compatible recipes

Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.

Record

Status
candidate
Source
localmaxxing
Engine
vllm
Engine version
0.20.1-local
Accelerators
4
Tensor parallel
4
Context tokens
2,048
Max concurrency
1
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/MiniMaxAI/MiniMax-M2.7
Repository
MiniMaxAI/MiniMax-M2.7
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

Source
https://www.localmaxxing.com/en/runs/cmp4f31dh000amz01uqa329e6

Observed source tokens

Mechanical split of the source command. Unverified against the engine CLI. The registry does not offer Run for this recipe.

  1. vllm
  2. bench
  3. throughput
  4. --backend
  5. vllm
  6. --async-engine
  7. --block-size
  8. 256
  9. --no-enable-prefix-caching
  10. --model
  11. /mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround
  12. --tokenizer
  13. /mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround
  14. --trust-remote-code
  15. --dtype
  16. float16
  17. --tensor-parallel-size
  18. 4
  19. --distributed-executor-backend
  20. mp
  21. --max-model-len
  22. 2048
  23. --max-num-batched-tokens
  24. 512
  25. --max-num-seqs
  26. 1
  27. --dataset-name
  28. random
  29. --random-input-len
  30. 512
  31. --random-output-len
  32. 1536
  33. --random-range-ratio
  34. 0
  35. --num-prompts
  36. 1
  37. --disable-log-stats
  38. --compilation-config
  39. {"use_inductor_graph_partition":true,"compile_sizes":[1],"cudagraph_mode":"PIECEWISE"}
FlagValue
--backendvllm
--block-size256
--model/mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround
--tokenizer/mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround
--dtypefloat16
--tensor-parallel-size4
--distributed-executor-backendmp
--max-model-len2048
--max-num-batched-tokens512
--max-num-seqs1
--dataset-namerandom
--random-input-len512
--random-output-len1536
--random-range-ratio0
--num-prompts1
--compilation-config{"use_inductor_graph_partition":true,"compile_sizes":[1],"cudagraph_mode":"PIECEWISE"}

Environment

VariableValue
CCL_TOPO_P2P_ACCESS1
VLLM_CACHE_ROOT/mnt/fast-ai/vllm-cache-exp/minimax-xpugraph-attndelay-block256-mbt512-noprefix-20260513T171301Z
VLLM_MINIMAX_M2_ATTN_DELAY_ALLREDUCE1
VLLM_XPU_ENABLE_XPU_GRAPH1
VLLM_XPU_FORCE_GRAPH_WITH_COMM1
VLLM_XPU_GRAPH_NOOP_COMM_CAPTURE1
VLLM_XPU_USE_LLM_SCALER_MOE1

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
12,048—73.3—observedminimax-m2-7-int4-autoround-w4a16-intel-arc-pro-b70-32gb-vllm-tp4-uqa329e6-sweep

Remaining fields

Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.

hardware count
4
hardware id
intel-arc-pro-b70-32gb
id
minimax-m2-7-int4-autoround-w4a16-intel-arc-pro-b70-32gb-vllm-tp4-uqa329e6
model instance id
minimaxai-minimax-m2-7--int4-autoround-w4a16
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
minimax-m2-7-int4-autoround-w4a16-intel-arc-pro-b70-32gb-vllm-tp4-uqa329e6-sweep
status
candidate

capabilities

engine

name
vllm
version
0.20.1-local

serving

max concurrency
1
max context tokens
2,048
tensor parallel
4
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
Intel Arc Pro B70
localmaxxing · notes
MiniMax M2.7 AutoRound W4A16 on 4x Intel Arc Pro B70 via vLLM/XPU TP4. Local weights are Lasimeri/MiniMax-M2.7-int4-AutoRound submitted under base MiniMaxAI/MiniMax-M2.7. Same recipe as prior 73.23 run, now repeated at 73.31 output tok/s and 97.74 total tok/s at p512/n1536. This is a variance-confirmed measured high, not a new quality-changing optimization. No model weight, quantization, router precision, expert routing, KV dtype, sampler, speculative decoding, or power-limit changes. Q/K RMS variance allreduces remain preserved. Total tok/s includes the 512-token prompt plus 1536 decoded tokens; no isolated prefill-only metric is emitted by this vLLM bench mode.
localmaxxing · observed command
VLLM_MINIMAX_M2_ATTN_DELAY_ALLREDUCE=1 VLLM_XPU_ENABLE_XPU_GRAPH=1 VLLM_XPU_FORCE_GRAPH_WITH_COMM=1 VLLM_XPU_GRAPH_NOOP_COMM_CAPTURE=1 VLLM_XPU_USE_LLM_SCALER_MOE=1 CCL_TOPO_P2P_ACCESS=1 VLLM_CACHE_ROOT=/mnt/fast-ai/vllm-cache-exp/minimax-xpugraph-attndelay-block256-mbt512-noprefix-20260513T171301Z vllm bench throughput --backend vllm --async-engine --block-size 256 --no-enable-prefix-caching --model /mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround --tokenizer /mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround --trust-remote-code --dtype float16 --tensor-parallel-size 4 --distributed-executor-backend mp --max-model-len 2048 --max-num-batched-tokens 512 --max-num-seqs 1 --dataset-name random --random-input-len 512 --random-output-len 1536 --random-range-ratio 0 --num-prompts 1 --disable-log-stats --compilation-config '{"use_inductor_graph_partition":true,"compile_sizes":[1],"cudagraph_mode":"PIECEWISE"}'
localmaxxing · run id
cmp4f31dh000amz01uqa329e6
localmaxxing · tokenized · arguments
vllm, bench, throughput, --backend, vllm, --async-engine, --block-size, 256, --no-enable-prefix-caching, --model, /mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround, --tokenizer, /mnt/fast-ai/llm-models/minimax-m2.7-int4-autoround, --trust-remote-code, --dtype, float16, --tensor-parallel-size, 4, --distributed-executor-backend, mp, --max-model-len, 2048, --max-num-batched-tokens, 512, --max-num-seqs, 1, --dataset-name, random, --random-input-len, 512, --random-output-len, 1536, --random-range-ratio, 0, --num-prompts, 1, --disable-log-stats, --compilation-config, {"use_inductor_graph_partition":true,"compile_sizes":[1],"cudagraph_mode":"PIECEWISE"}
localmaxxing · tokenized · environment · CCL TOPO P2P ACCESS
1
localmaxxing · tokenized · environment · VLLM MINIMAX M2 ATTN DELAY ALLREDUCE
1
localmaxxing · tokenized · environment · VLLM XPU ENABLE XPU GRAPH
1
localmaxxing · tokenized · environment · VLLM XPU FORCE GRAPH WITH COMM
1
localmaxxing · tokenized · environment · VLLM XPU GRAPH NOOP COMM CAPTURE
1
localmaxxing · tokenized · environment · VLLM XPU USE LLM SCALER MOE
1
localmaxxing · tokenized · fidelity
faithful

provenance

captured at
2026-08-30T09:10:02Z

sources

captured atkindurl
2026-08-30T09:10:02Znormalized-recipewww.localmaxxing.com/en/runs/cmp4f31dh000amz01uqa329e6 ↗