NOUMENOpen app
NOUMEN

MiniMax · Open weights

MiniMax M1

Added Jan 8, 2025Open weights

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

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.

Input
Text
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.14
Output
$1.33

Reference prices as published by the model provider. API example

Specifications

Context window
1M
Max output
131.1K

Side by side

MiniMax M1 compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

MiniMax M1MiniMaxThis modelMiniMax M3MiniMaxRemove
ProviderMiniMaxMiniMax
Discovery tierOpen weightsOpen weights
AddedJan 8, 2025Jun 1, 2026
Limits
Context window1MBest512K
Max output131.07KBest80K
Average output—491 tokens
Pricing, USD per 1M tokens
Input$0.14Best$0.23
Output$1.33$0.96Best
Cache read—$0.05
Performance
Throughput—202.4 tok/s
Time to first token—2.8 s
Request success (24h)—99.9%
Benchmark indexes
Intelligence Index10.029.2Best
Coding Index—58.6
Math Index13.7—
Capabilities
ToolsNoYes
VisionNoYes
Structured outputNoYes
Reasoning effort——

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • MiniMax M1
  • MiniMax M3

Intelligence Index

  • MiniMax M110.0
  • MiniMax M329.2

MiniMax M1: Better than 21% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • MiniMax M1: —
    • MiniMax M3: 58.6

  • Math Index

    Composite of the math evaluations

    • MiniMax M1: 13.7
    • MiniMax M3: —

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • MiniMax M1: —
    • MiniMax M3: 21.3%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • MiniMax M1: —
    • MiniMax M3: 4.4%

  • Harvey LAB-AAv1.1

    Legal tasks passed in full, with no material hallucination

    • MiniMax M1: —
    • MiniMax M3: 0.0%

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • MiniMax M1: —
    • MiniMax M3: 1,089 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • MiniMax M1: —
    • MiniMax M3: 1,243 Elo

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • MiniMax M1: —
    • MiniMax M3: 9.8%

  • AA-LCRv1.1

    Reasoning across long documents

    • MiniMax M1: —
    • MiniMax M3: 83.0%

  • MLCR-AA

    Long-context reasoning over medical records

    • MiniMax M1: —
    • MiniMax M3: 4.4%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • MiniMax M1: 7.8%
    • MiniMax M3: 39.0%Top

  • GPQA Diamond

    Graduate-level science questions

    • MiniMax M1: 68.2%
    • MiniMax M3: 92.9%Top

  • MMLU-Pro

    Multi-subject knowledge and reasoning

    • MiniMax M1: 80.8%
    • MiniMax M3: —

  • IFBench

    Following precise output instructions

    • MiniMax M1: 41.2%
    • MiniMax M3: 82.9%Top

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • MiniMax M1: —
    • MiniMax M3: 2.0%

  • SciCode

    Scientific computing in Python

    • MiniMax M1: —
    • MiniMax M3: 47.1%

  • LiveCodeBench

    Recent competitive programming problems

    • MiniMax M1: 65.7%
    • MiniMax M3: —

Math

  • AIME 2025

    Competition mathematics, 2025 problems

    • MiniMax M1: 13.7%
    • MiniMax M3: —

  • AIME

    Competition mathematics

    • MiniMax M1: 81.3%
    • MiniMax M3: —

  • MATH-500

    Competition math problems

    • MiniMax M1: 97.2%
    • MiniMax M3: —

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • MiniMax M1: 2.3%
    • MiniMax M3: 42.4%Top

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • MiniMax M1: 31.6%
    • MiniMax M3: 88.9%Top

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • MiniMax M1: —
    • MiniMax M3: 83.0%

Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "MiniMax-M1",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="MiniMax-M1",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

MiniMax

Other MiniMax models, starting with the same family.