Features
1M token context window with MSA sparse attention
Native multimodal understanding (text, code, images, video, audio, music)
Leading coding capabilities (SWE-Bench Verified 80.2%)
Autonomous agent team formation via MiniMax Code
Persistent user memory and skill customization
Multi-modal generation: text, video (Hailuo 2.3), speech (Speech 2.8), music (Music 2.6)
Token Plan subscription model with cost advantage
MaxProof reinforcement learning for math proofs
Desktop coding agent (MiniMax Code)
API access with Token Plan
Open-source model weights (M3 released on Hugging Face)
Supports 200+ countries and regions
Local model execution for offline inference
Connect online APIs: OpenAI (ChatGPT), Anthropic (Claude), Google (Gemini), Meta (Llama), Mistral AI, Alibaba (Qwen),
123+ HuggingFace models available
Memory feature (coming soon) for context persistence
Minimalist UI with dark mode
Open source under MIT license on GitHub
Cross-platform desktop app (Mac, Windows, Linux)
Version 0.8.3 with bug fixes and performance improvements