# MiniMax-M1-80k

> We introduce MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model. MiniMax-M1 is powered by a hybrid Mixture-of-Experts (MoE) architecture combined with a lightning attention mechanism. The model is developed based on our previous MiniMax-Text-01 model, which contains a total of 456 billion parameters with 45.9 billion parameters activated per token.

Provider: MiniMax · Type: LLM · State: Ready

model path: `accounts/fireworks/models/minimax-m1-80k`

[Deploy on demand](https://app.fireworks.ai/dashboard/deployments/create?baseModel=accounts/fireworks/models/minimax-m1-80k)

We introduce MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model. MiniMax-M1 is powered by a hybrid Mixture-of-Experts (MoE) architecture combined with a lightning attention mechanism. The model is developed based on our previous MiniMax-Text-01 model, which contains a total of 456 billion parameters with 45.9 billion parameters activated per token.

## MiniMax-M1-80k API Features

- **On-demand Deployment**: On-demand deployments allow you to use MiniMax-M1-80k on dedicated GPUs with Fireworks' high-performance serving stack with high reliability and no rate limits. · [Docs](https://docs.fireworks.ai/guides/ondemand-deployments)

## Metadata

- **State**: Ready
- **Created on**: 19 JUN, 2025
- **Kind**: Base model
- **Provider**: MiniMax
- **Huggingface**: [MiniMaxAI/MiniMax-M1-80k/tree/main](https://huggingface.co/MiniMaxAI/MiniMax-M1-80k/tree/main)

## Specifications

- **Calibrated**: No
- **Mixture-of-Experts**: No

## Supporting Functionality

- **Fine-tuning**: Not Supported
- **Serverless**: Not Supported
- **Context Length**: N/A
- **Function Calling**: Not Supported
- **Embeddings**: Not Supported
- **Rerankers**: Not Supported
- **Supporting image input**: Not Supported
