MiniMax M3

MiniMax-M3 is a multimodal foundation model from MiniMax. It accepts text, image, and video inputs and produces text output, with a 1M-token context window, and is suited for long-horizon agentic work, coding, and tool use. It is built on MiniMax Sparse Attention (MSA), which replaces full attention with KV-block selection to cut per-token compute at long context, roughly 1/20 the cost of the previous generation at 1M tokens, with substantially faster prefill and decode while retaining quality across most tasks. Trained as a native multimodal model on interleaved data and tuned for multi-turn, production-like collaboration via an interactive user-simulator framework, it is oriented toward sustained, multi-step tasks rather than single-turn execution.

MiniMax M2.5

MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diverse range of complex real-world digital working environments, M2.5 builds upon the coding expertise of M2.1 to extend into general office work, reaching fluency in generating and operating Word, Excel, and Powerpoint files, context switching between diverse software environments, and working across different agent and human teams. Scoring 80.2% on SWE-Bench Verified, 51.3% on Multi-SWE-Bench, and 76.3% on BrowseComp, M2.5 is also more token efficient than previous generations, having been trained to optimize its actions and output through planning.

MiniMax M1

MiniMax M1 is a large-scale, open-weight reasoning model designed for extended context and high-efficiency inference. It leverages a hybrid Mixture-of-Experts (MoE) architecture paired with a custom "lightning attention" mechanism, allowing it to process long sequences (up to 1 million tokens) while maintaining competitive FLOP efficiency. With 456 billion total parameters and 45.9B active per token, this variant is optimized for complex, multi-step reasoning tasks.

MiniMax M2-her

MiniMax M2-her is a dialogue-first large language model built for immersive roleplay, character-driven chat, and expressive multi-turn conversations. Designed to stay consistent in tone and personality, it supports rich message roles and can learn from example dialogue to better match the style and pacing of your scenario.

MiniMax: MiniMax M2.1

MiniMax: MiniMax M2.1 model integrated via automation on 2025-12-23

MiniMax M2

MiniMax M2 is a high-efficiency 10B activated parameter model optimized for coding agents, compile-run-fix loops, and long-horizon reasoning. It balances responsiveness with strong SWE-Bench and Terminal-Bench results, excels at code generation, planning, and tool use, and preserves reasoning continuity across multi-step tasks.

MiniMax M3

MiniMax-M3 is a multimodal foundation model from MiniMax. It accepts text, image, and video inputs and produces text output, with a 1M-token context window, and is suited for long-horizon agentic work, coding, and tool use. It is built on MiniMax Sparse Attention (MSA), which replaces full attention with KV-block selection to cut per-token compute at long context, roughly 1/20 the cost of the previous generation at 1M tokens, with substantially faster prefill and decode while retaining quality across most tasks. Trained as a native multimodal model on interleaved data and tuned for multi-turn, production-like collaboration via an interactive user-simulator framework, it is oriented toward sustained, multi-step tasks rather than single-turn execution.

MiniMax M2.5

MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diverse range of complex real-world digital working environments, M2.5 builds upon the coding expertise of M2.1 to extend into general office work, reaching fluency in generating and operating Word, Excel, and Powerpoint files, context switching between diverse software environments, and working across different agent and human teams. Scoring 80.2% on SWE-Bench Verified, 51.3% on Multi-SWE-Bench, and 76.3% on BrowseComp, M2.5 is also more token efficient than previous generations, having been trained to optimize its actions and output through planning.

MiniMax M1

MiniMax M1 is a large-scale, open-weight reasoning model designed for extended context and high-efficiency inference. It leverages a hybrid Mixture-of-Experts (MoE) architecture paired with a custom "lightning attention" mechanism, allowing it to process long sequences (up to 1 million tokens) while maintaining competitive FLOP efficiency. With 456 billion total parameters and 45.9B active per token, this variant is optimized for complex, multi-step reasoning tasks.

MiniMax M2-her

MiniMax M2-her is a dialogue-first large language model built for immersive roleplay, character-driven chat, and expressive multi-turn conversations. Designed to stay consistent in tone and personality, it supports rich message roles and can learn from example dialogue to better match the style and pacing of your scenario.

MiniMax: MiniMax M2.1

MiniMax: MiniMax M2.1 model integrated via automation on 2025-12-23

MiniMax M2

MiniMax M2 is a high-efficiency 10B activated parameter model optimized for coding agents, compile-run-fix loops, and long-horizon reasoning. It balances responsiveness with strong SWE-Bench and Terminal-Bench results, excels at code generation, planning, and tool use, and preserves reasoning continuity across multi-step tasks.