vLLM Announces Day-0 Support for MiniMax M3.1 Flash with NVFP4 Gains
vLLM provides day-0 serving support for MiniMax M3.1 Flash, with native NVFP4 cutting checkpoint size by 44.6% and boosting 256K-context concurrency 2.7x, reaching 191.5k tok/s per GPU on Vera Rubin NVL72 alongside strong benchmark results.
2026-10-23 ~ 2026-10-23 · 4 related posts
- Episode 1: MiniMax-M3.1-Flash scores Intelligence Index 42, ranks 4th but shows 93% hallucination rate in AA evals(2026-10-19, 9 posts)
- Episode 2: MiniMax Open-Sources M3.1-Flash: 428B Multimodal MoE with 1M Context(2026-10-23, 8 posts)
- Episode 3: vLLM Announces Day-0 Support for MiniMax M3.1 Flash with NVFP4 Gains(2026-10-23, 4 posts)
- M3.1 Flash hits 191.5k tok/s per GPU on Vera Rubin NVL72, leads open models on agent benchmarks — vllm_project · 2026-10-23
- MiniMax M3.1 Flash benchmarks: 91.0 BabyVision, 48% HLE, top BankerToolBench score — vllm_project · 2026-10-23
- vLLM details M3.1 Flash serving: NVFP4 cuts weights 44.6%, boosts capacity 2.7x — vllm_project · 2026-10-23
- vLLM ships Day-0 serving for MiniMax M3.1 Flash: NVFP4 cuts KV memory 43.75%, 2.7× concurrency — vllm_project · 2026-10-23