MiniMax M3 scores 59% on SWE-Bench Pro, above GPT-5.5 and Gemini 3.1 Pro, with open weights due on Hugging Face
Shanghai's MiniMax released M3, an open-weight model that hits 59.0% on SWE-Bench Pro — clearing GPT-5.5 and Gemini 3.1 Pro — and processes a 1M-token context window at one-twentieth the per-token compute of prior generations, with 9× faster prefill and 15× faster decoding. Open weights arrive on Hugging Face within ten days, putting frontier-grade long-context coding within reach for self-hosted deployments — a competitive signal timed to MiniMax's Shanghai Star Market listing push.
Source: scmp.com ↗
the model's redesigned architecture reduced computational requirements to as little as one-twentieth of previous levels, slashing inference costs while boosting response speeds.
MiniMax
Why this matters
- → MiniMax M3 reduces inference costs to one-twentieth of previous levels while handling 1M-token contexts, enabling cost-effective long-context coding at scale.
- → Open weights on Hugging Face within ten days democratize frontier-grade coding AI for self-hosted deployments.
- → Outperforming GPT-5.5 and Gemini 3.1 Pro on SWE-Bench Pro signals competitive pressure in the coding-agent market.
China's coding AI surge