MiniMax M3.1 Flash, a multimodal coding-focused LLM with a 1M token context window, is now available for free on OpenRouter. Previously only accessible through MiniMax's proprietary tools, this model supports tool use and agentic workflows, with a unique feature that keeps reasoning always on, adjustable by effort level. Early user reports suggest decode speeds between 90-110 tokens/sec, though official benchmarks are pending. AI
IMPACT Provides a powerful, free multimodal coding assistant with a large context window, potentially accelerating agent development and coding assistance.
RANK_REASON New model release from a frontier lab (MiniMax) with significant capability (1M context, multimodal) and pricing change (free). [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- DeepSeek V4.1 Flash
- MiniMax
- MiniMax Code
- MiniMax M3
- MiniMax M3.1 Flash
- MiniMax Sparse Attention
- OpenRouter
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →