Every model.One place.
Explore the models available through Noumen. Compare capabilities, context windows, and reference pricing, then see which models people use most and how they score on public benchmarks.
Just landed
Recently added
The newest models available through Noumen, ready to use in the app and the API.
- FrontierNew
- FrontierNew
- Open weightsNew
- EfficientNew
MiniMax M2-herMiniMaxCapabilities: Text generationContext65.5KInput / 1M$0.302Output / 1M$1.207Show model details
Dialogue-focused MiniMax model optimized for role-playing and multi-turn conversations with rich role settings.
- Model ID
minimax/minimax-m2-her- Discovery tier
- Balanced
- Context window
- 65,532 tokens
MiniMax M1 80KMiniMaxCapabilities: Text generationContext1MInput / 1M$0.6052Output / 1M$2.4225Show model details
MiniMax-M1 with 80K thinking budget. Enhanced version of the hybrid MoE reasoning model with double the thinking capacity. Ideal for extremely complex software engineering, tool use, and long context tasks.
- Model ID
MiniMaxAI/MiniMax-M1-80k- Discovery tier
- Balanced
- Context window
- 1,000,000 tokens
MiniMax M3.1 FlashMiniMaxCapabilities: Reasoning, Tools, Vision, Video inputContext1MInput / 1M$0.15Output / 1M$0.6Show model details
MiniMax M3.1 Flash is a multimodal model for coding, tool use, and long-context work. It accepts text, images, and video and produces text, with a 1,000,000-token context window and up to 524,288 output tokens within the remaining context. Thinking is always enabled, with low, medium, high, xhigh, and max effort levels; max is the default. Thinking tokens count toward the output limit. Temperature is fixed at 1 and top-p at 0.95; custom values have no effect.
- Model ID
minimax/minimax-m3.1-flash- Discovery tier
- Efficient
- Context window
- 1,048,576 tokens
MiniMax 01MiniMaxCapabilities: PDF inputContext1MInput / 1M$0.1394Output / 1M$1.122Show model details
MiniMax's flagship model with a 1M token context window
- Model ID
minimax/minimax-01- Discovery tier
- Efficient
- Context window
- 1,000,192 tokens
MiniMax M3MiniMaxCapabilities: Tools, Vision, Structured outputContext512KInput / 1M$0.23Output / 1M$0.96Show model details
MiniMax M3 is the non-thinking route for MiniMax's open-weights frontier model, built for coding, agent workflows, tool use, and multimodal understanding from step zero. It keeps native thinking disabled for faster direct answers. MiniMax reports 59.0% on SWE-Bench Pro and 66.0% on Terminal Bench 2.1, with Sparse Attention designed to scale context to 1M.
- Model ID
minimax/minimax-m3- Discovery tier
- Open weights
- Context window
- 512,000 tokens
MiniMax M3 ThinkingMiniMaxCapabilities: Reasoning, Tools, Vision, Structured outputContext512KInput / 1M$0.23Output / 1M$0.96Show model details
MiniMax M3 Thinking is the adaptive-thinking version of MiniMax's open-weights frontier model for coding, agent workflows, tool use, long-context tasks, and native multimodal understanding. MiniMax reports 59.0% on SWE-Bench Pro and 66.0% on Terminal Bench 2.1, with Sparse Attention designed to scale context to 1M.
- Model ID
minimax/minimax-m3:thinking- Discovery tier
- Open weights
- Context window
- 512,000 tokens
MiniMax LatestMiniMaxCapabilities: Reasoning, Tools, Vision, Structured outputContext512KInput / 1M$0.23Output / 1M$0.96Show model details
Compatibility alias for MiniMax M3 Thinking, retaining its existing request format and structured-output support. The newer MiniMax M3.1 Flash is available separately as minimax/minimax-m3.1-flash and has different request parameter support.
- Model ID
minimax/minimax-latest- Discovery tier
- Open weights
- Context window
- 512,000 tokens
MiniMax M2.7MiniMaxCapabilities: Reasoning, Tools, Structured outputContext204.8KInput / 1M$0.21Output / 1M$0.84Show model details
MiniMax M2.7 is the first model deeply involved in iterating on its own training. It excels in real-world software engineering (SWE-Pro 56.22%), end-to-end project delivery (VIBE-Pro 55.6%), and complex office workflows with strong tool-use compliance and agentic capabilities.
- Model ID
minimax/minimax-m2.7- Discovery tier
- Open weights
- Context window
- 204,800 tokens
MiniMax M2.7 TurboMiniMaxCapabilities: Reasoning, Tools, Structured outputContext204.8KInput / 1M$0.6Output / 1M$2.4Show model details
MiniMax M2.7 Turbo is the highspeed and higher priced route for M2.7.
- Model ID
minimax/minimax-m2.7-turbo- Discovery tier
- Open weights
- Context window
- 204,800 tokens
MiniMax M2.5MiniMaxCapabilities: Reasoning, Tools, Structured outputContext204.8KInput / 1M$0.27Output / 1M$1.08Show model details
MiniMax M2.5 is a productivity-focused flagship model that builds on M2.1 with stronger coding and real-world office workflow performance (Word, Excel, PowerPoint), plus better tool-use planning and token efficiency.
- Model ID
minimax/minimax-m2.5- Discovery tier
- Open weights
- Context window
- 204,800 tokens
MiniMax M2.1MiniMaxCapabilities: Reasoning, Tools, Structured outputContext200KInput / 1M$0.3Output / 1M$1.2Show model details
MiniMax M2.1 builds on M2 with enhanced context understanding and improved complex tool use. 230B parameter MoE model (10B active) optimized for agentic workflows and long-horizon tasks.
- Model ID
minimax/minimax-m2.1- Discovery tier
- Open weights
- Context window
- 200,000 tokens
MiniMax M2MiniMaxCapabilities: ReasoningContext200KInput / 1M$0.17Output / 1M$1.53Show model details
MiniMax M2 offers enhanced reasoning and strong general performance. Optimized for coding and agentic workflows.
- Model ID
minimax/minimax-m2- Discovery tier
- Open weights
- Context window
- 200,000 tokens
MiniMax M1MiniMaxCapabilities: Text generationContext1MInput / 1M$0.1394Output / 1M$1.3328Show model details
MiniMax-M1 is a hybrid MoE reasoning model with 40K thinking budget. World's first open-weight, large-scale hybrid-attention model with lightning attention for efficient test-time compute scaling. Excels at complex tasks requiring extensive reasoning.
- Model ID
MiniMax-M1- Discovery tier
- Open weights
- Context window
- 1,000,000 tokens
Reference pricing in USD per 1M tokens, as published by model providers. Discovery tiers describe model metadata, rather than benchmark rankings.