MiniMax-M2.5
By MiniMaxAI. 228.7 billion parameters, a context window of 196,608 tokens and the licence other.
Facts
- Released
- 2026-02-12 the day the repository was first published on Hugging Face Hugging Face, read
- Licence
- other the licence the model card declares Hugging Face, read
- Open weights
- Yes the weights are published in this Hugging Face repository Hugging Face, read
- Parameters
- 228.7 billion counted from the safetensors weight files Hugging Face, read
- Active parameters
- 11.0 billion estimate estimated from the expert counts and sizes in the configuration model configuration, read
- Knowledge cutoff
- Unknown
- Context window
- 196,608 tokens the maximum position embeddings in the model configuration model configuration, read
- Longest output
- Unknown
- Tool calling
- Unknown
- Structured output
- Unknown
- Reasoning controls
- Unknown
- Inputs and outputs
- text in, text out Hugging Face, read
- Good for
- No source names a use.
Prices
| Provider | Input | Output | Context | As of | Source |
|---|---|---|---|---|---|
| Deep Infra | 0.150 USD | 1.15 USD | 196,608 | older than 30 days; check the provider | models.dev, read |
| Baseten | 0.300 USD | 1.20 USD | 204,000 | older than 30 days; check the provider | models.dev, read |
| CloudFerro Sherlock | 0.300 USD | 1.20 USD | 196,000 | older than 30 days; check the provider | models.dev, read |
| CommandCode | 0.300 USD | 1.20 USD | 200,000 | older than 30 days; check the provider | models.dev, read |
| Friendli | 0.300 USD | 1.20 USD | 196,608 | older than 30 days; check the provider | models.dev, read |
| Hugging Face | 0.300 USD | 1.20 USD | 204,800 | older than 30 days; check the provider | models.dev, read |
| Meganova | 0.300 USD | 1.20 USD | 204,800 | older than 30 days; check the provider | models.dev, read |
| Nebius Token Factory | 0.300 USD | 1.20 USD | 196,608 | LiteLLM, read | |
| SiliconFlow | 0.300 USD | 1.20 USD | 197,000 | older than 30 days; check the provider | models.dev, read |
| Together AI | 0.300 USD | 1.20 USD | 204,800 | older than 30 days; check the provider | models.dev, read |
| tensormesh | 0.300 USD | 1.20 USD | 196,608 | LiteLLM, read |
Run it on your own hardware
Weights are the sizes of the files a source lists, or an estimate from the parameter count where none does. Memory is an estimate: weights plus KV cache plus 512 MiB and 5 percent of the weights for runtime buffers. Check it against your hardware.
| Quantization | Weights | Memory at 8,192 tokens | Memory at 32,768 tokens |
|---|---|---|---|
| UD-TQ1_0 | 51.9 GiB | 57.0 GiB | 62.8 GiB |
| UD-IQ1_S | 58.9 GiB | 64.2 GiB | 70.1 GiB |
| UD-IQ1_M | 63.7 GiB | 69.4 GiB | 75.2 GiB |
| UD-IQ2_XXS | 69.0 GiB | 74.9 GiB | 80.7 GiB |
| UD-IQ2_M | 72.8 GiB | 78.9 GiB | 84.7 GiB |
| Q2_K | 77.6 GiB | 83.9 GiB | 89.7 GiB |
| Q2_K_L | 77.7 GiB | 84.0 GiB | 89.9 GiB |
| UD-Q2_K_XL | 80.0 GiB | 86.5 GiB | 92.3 GiB |
| UD-IQ3_XXS | 86.9 GiB | 93.7 GiB | 99.5 GiB |
| Q3_K_S | 91.9 GiB | 98.9 GiB | 105 GiB |
| UD-Q3_K_XL | 94.3 GiB | 101 GiB | 107 GiB |
| Q3_K_M | 102 GiB | 109 GiB | 115 GiB |
| IQ4_XS | 114 GiB | 122 GiB | 127 GiB |
| MXFP4_MOE | 115 GiB | 123 GiB | 129 GiB |
| IQ4_NL | 120 GiB | 129 GiB | 134 GiB |
| Q4_0 | 121 GiB | 129 GiB | 135 GiB |
| Q4_K_S | 121 GiB | 130 GiB | 135 GiB |
| UD-Q4_K_XL | 122 GiB | 131 GiB | 137 GiB |
| Q4_K_M | 129 GiB | 138 GiB | 144 GiB |
| Q4_1 | 133 GiB | 142 GiB | 148 GiB |
| Q5_K_S | 147 GiB | 156 GiB | 162 GiB |
| UD-Q5_K_XL | 151 GiB | 161 GiB | 166 GiB |
| Q5_K_M | 151 GiB | 161 GiB | 167 GiB |
| Q6_K | 175 GiB | 186 GiB | 192 GiB |
| UD-Q6_K_XL | 181 GiB | 192 GiB | 198 GiB |
| Q8_0 | 226 GiB | 240 GiB | 246 GiB |
| UD-Q8_K_XL | 243 GiB | 258 GiB | 264 GiB |
| BF16 | 426 GiB | 450 GiB | 456 GiB |
Start it with a local runtime
- Ollama
ollama run hf.co/unsloth/MiniMax-M2.5-GGUF:Q4_0 - LM Studio
lms get unsloth/MiniMax-M2.5-GGUF lms server start - llama.cpp server
llama-server -hf unsloth/MiniMax-M2.5-GGUF:Q4_0 --jinja - vLLM
vllm serve MiniMaxAI/MiniMax-M2.5 - SGLang
sglang serve --model-path MiniMaxAI/MiniMax-M2.5 --port 30000
Use it from your harness
Set up for Baseten with the model MiniMaxAI/MiniMax-M2.5. Each endpoint page has the same setup for its own address.
OpenCode
Put this in opencode.json in your project folder:
{
"$schema": "https://opencode.ai/config.json",
"provider": {
"baseten": {
"npm": "@ai-sdk/openai-compatible",
"name": "Baseten",
"options": {
"baseURL": "https://inference.baseten.co/v1",
"apiKey": "{env:BASETEN_API_KEY}"
},
"models": {
"MiniMaxAI/MiniMax-M2.5": {
"name": "MiniMax-M2.5"
}
}
}
}
}- OpenCode reads any OpenAI-compatible address through the @ai-sdk/openai-compatible package, and an address that speaks the Responses API through @ai-sdk/openai.
From OpenCode documentation, read .
Pi
Put this in ~/.pi/agent/models.json:
{
"providers": {
"baseten": {
"baseUrl": "https://inference.baseten.co/v1",
"api": "openai-completions",
"apiKey": "$BASETEN_API_KEY",
"models": [
{
"id": "MiniMaxAI/MiniMax-M2.5"
}
]
}
}
}- The apiKey field can name an environment variable as $NAME.
From Pi documentation, read .
Codex
Codex speaks only the Responses API, and Baseten documents no Responses address. A gateway that offers one can sit in between.
- Codex speaks the Responses API only: responses is the one supported wire API of a custom provider. Ollama and LM Studio are built in and start with --oss.
From Codex documentation, read .
Claude Code
Claude Code sends Anthropic Messages requests, and Baseten documents no such address. A gateway that translates to that API can sit in between.
- Claude Code sends Anthropic Messages requests to ANTHROPIC_BASE_URL. Anthropic says it does not support routing Claude Code to models other than Claude through any gateway, so some features may not work with another model.
From Claude Code documentation, read .
Other ways to reach it
- CloudFerro Sherlock, model
MiniMaxAI/MiniMax-M2.5 - CommandCode, model
MiniMaxAI/MiniMax-M2.5 - DeepInfra, model
MiniMaxAI/MiniMax-M2.5 - Friendli, model
MiniMaxAI/MiniMax-M2.5 - Hugging Face, model
MiniMaxAI/MiniMax-M2.5 - Meganova, model
MiniMaxAI/MiniMax-M2.5 - Nebius Token Factory, model
MiniMaxAI/MiniMax-M2.5 - SiliconFlow, model
MiniMaxAI/MiniMax-M2.5 - Together AI, model
MiniMaxAI/MiniMax-M2.5 - Ollama, model
hf.co/unsloth/MiniMax-M2.5-GGUF:Q4_0
Published results
These are results other people published. Baltor did not run them and does not rank models by them.
- Results the maker published on the model card published by MiniMaxAI Hugging Face, read
Two more publishers are linked, never copied, because their terms do not allow it:
- Artificial Analysis publishes its own intelligence, speed and price measurements.
- OpenRouter lists current prices and uptime for each provider it routes to.
Sources of this page
Paid links
No link in this directory is a paid link or an ad, and no listing is paid for. The order and the contents of every list come from the sources named on this page.