MiniMax has released its M3.1 Flash model, which shows significant improvements over its predecessor, scoring 42 on the Artificial Analysis Intelligence Index. This new model is also notably cheaper, operating at half the cost of the previous version. On the DeepSWE v1.1 coding benchmark, M3.1 Flash achieved 73.7 percent for under $0.30 per task, matching the performance and price point of Deepseek V4.1 Flash 1017. AI
IMPACT This release intensifies the price competition in the AI market, potentially signaling a lack of a price floor for reasoning models in China.
RANK_REASON Frontier-lab model release with performance and pricing details. [lever_c_demoted from frontier_release: ic=2 ai=1.0]
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →