MiniMax-M3.1-Flash / README.md
tokenfactorypacketai's picture
Update README.md
771b6ba verified
|
Raw History Blame Contribute Delete
5.48 kB
metadata
license: other
license_name: minimax-community
license_link: https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash
pipeline_tag: image-text-to-text
base_model: MiniMaxAI/MiniMax-M3.1-Flash
tags:
  - minimax
  - minimax-m3.1
  - multimodal
  - mixture-of-experts
  - moe
  - agent
  - coding
  - computer-use
  - long-context
  - conversational
  - openai-compatible
  - serverless-inference
  - llm-api
  - token-factory
  - packet-ai

MiniMax-M3.1-Flash — Serverless API on Token Factory by packet.ai

Run MiniMax-M3.1-Flash through a serverless, OpenAI-compatible API with pay-per-token billing. Token Factory by packet.ai serves open-weight models behind an OpenAI-compatible endpoint, so you can switch your existing OpenAI SDK code by changing only the base URL and API key.

This card describes the hosted API for MiniMaxAI/MiniMax-M3.1-Flash, MiniMax's native multimodal mixture-of-experts model — an agentic model for coding and computer use.

Why use MiniMax-M3.1-Flash on Token Factory

  • OpenAI-compatible endpoint — works with existing OpenAI SDKs.
  • Streaming, JSON mode and tool calling supported on chat completions.
  • Pay per token — input and output tokens are metered separately; nothing is charged while idle.
  • Free for one month at launch (till Dec 16th), then billed per 1M tokens. See current pricing.
  • Scale to dedicated GPUs — move the model to a dedicated GPU running vLLM while keeping the same endpoint shape.

Good for

MiniMax describes M3.1-Flash as built for:

  • Coding
  • Agentic workflows
  • Computer use
  • Cybersecurity

Model overview

Property Value
Model MiniMax-M3.1-Flash
Developer MiniMax
Type Native multimodal mixture-of-experts (MoE)
Builds on MiniMax-M3
Attention MiniMax Sparse Attention (MSA)
Context window Up to 1M tokens (usable length depends on serving configuration)
Input Text, image, video
Output Text
Reasoning effort levels low, medium, high, xhigh, max
Recommended sampling temperature = 1.0, top_p = 0.95
License MiniMax Community License

Benchmarks

Selected results reported by MiniMax for the official MiniMax-M3.1-Flash weights. See the original model card for the full table and comparisons.

Benchmark MiniMax-M3.1-Flash
DeepSWE v1.1 73.7
NL2Repo-Bench 61.2
Terminal Bench 4.0 38.1
OSWorld 2.0 61.7
Automation-Bench v1.0.6 62.8
BankerToolBench 76.2
CyberGym 82.2
LOCA-Bench 256K 60.3
BabyVision (with tools) 91.0
CharXiv RQ 82.5
HLE (without tools) 43.4

Quickstart

1. Get an API key

Sign up at dash.packet.ai and create an API key.

2. Call the API (Python, OpenAI SDK)

from openai import OpenAI
 
client = OpenAI(
    api_key="PACKET_API_KEY",
    base_url="https://api.packet.ai/v1",
)
 
response = client.chat.completions.create(
    model="MINIMAX_MODEL_ID_FROM_DASHBOARD",
    messages=[{"role": "user", "content": "Implement Hello World"}],
)
print(response.choices[0].message.content)

Replace PACKET_API_KEY with your API key. Full API documentation: dash.packet.ai/docs

Other models on Token Factory

Model Highlights
nvidia/DeepSeek-V4.1-Flash-NVFP4 Reasoning model, 1M-token context, image input
Qwen/Qwen2.5-1.5B-Instruct Small, fast model for classification and extraction, 32K context, 29+ languages

See the full, up-to-date list at packet.ai/token-factory.

Run it yourself

Prefer to self-host? Use the original weights from MiniMax: MiniMaxAI/MiniMax-M3.1-Flash. MiniMax recommends SGLang, vLLM or TokenSpeed for serving.

About packet.ai

packet.ai is a hosted·ai project operated by HOSTED AI Inc., 2880 Zanker Rd, Suite 203, San Jose, CA 95134. Alongside Token Factory, packet.ai offers GPUs (B200, RTX PRO 6000, A100, RTX 4090), dynamic and dedicated GPUs, and bare metal/clusters, billed by the minute.

License and attribution

MiniMax-M3.1-Flash is developed by MiniMax and released under the MiniMax Community License; see the original model card for the license terms. Token Factory provides hosted inference for this model. Use of the API is subject to packet.ai's Terms and Privacy Policy.

Citation

MiniMax Sparse Attention paper: arXiv:2606.13392