NOUMENOpen app
NOUMEN

MiniMax · Open weights

MiniMax Latest

Added May 3, 2026Open weights

Compare models

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

Loading models…

Use in the app

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.

  • Reasoning
  • Tools
  • Vision
  • Structured output
Input
Text, Images
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.23
Output
$0.96
Cache read
$0.05

Reference prices as published by the model provider. API example

Specifications

Context window
512K
Max output
80K

Side by side

MiniMax Latest 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 LatestMiniMaxThis modelGPT 6 LunaOpenAIRemove
ProviderMiniMaxOpenAI
Discovery tierOpen weightsEfficient
AddedMay 3, 2026Sep 22, 2026
Limits
Context window512K1.05MBest
Max output80K128KBest
Average output—120 tokens
Pricing, USD per 1M tokens
Input$0.23$0.10Best
Output$0.96$0.50Best
Cache read$0.05$0.01Best
Cache write—$0.13
Performance
Throughput170.7 tok/sBest68.2 tok/s
Time to first token1.2 sBest4.9 s
Request success (24h)99.9%Best99.8%
Benchmark indexes
Intelligence Index29.229.9Best
Coding Index58.6—
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesYes
Structured outputYesYes
PDF inputNoYes
Reasoning effort—none, low, medium, high, xhigh, max

Public evaluations

Benchmarks

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

  • MiniMax Latest
  • GPT 6 Luna

Intelligence Index

  • MiniMax Latest29.2
  • GPT 6 Luna29.9

MiniMax Latest: Better than 70% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • MiniMax Latest: 58.6
    • GPT 6 Luna: —

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • MiniMax Latest: 21.3%
    • GPT 6 Luna: 40.5%Top

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • MiniMax Latest: 4.4%
    • GPT 6 Luna: 10.2%Top

  • Harvey LAB-AAv1.1

    Legal tasks passed in full, with no material hallucination

    • MiniMax Latest: 0.0%
    • GPT 6 Luna: —

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • MiniMax Latest: 1,089 Elo
    • GPT 6 Luna: 1,097 EloTop

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • MiniMax Latest: 1,243 Elo
    • GPT 6 Luna: 1,263 EloTop

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • MiniMax Latest: 9.8%
    • GPT 6 Luna: 13.8%Top

  • AA-LCRv1.1

    Reasoning across long documents

    • MiniMax Latest: 83.0%Top
    • GPT 6 Luna: 78.3%

  • MLCR-AA

    Long-context reasoning over medical records

    • MiniMax Latest: 4.4%
    • GPT 6 Luna: 18.3%Top

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • MiniMax Latest: 39.0%Top
    • GPT 6 Luna: 28.3%

  • GPQA Diamond

    Graduate-level science questions

    • MiniMax Latest: 92.9%
    • GPT 6 Luna: —

  • IFBench

    Following precise output instructions

    • MiniMax Latest: 82.9%
    • GPT 6 Luna: —

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • MiniMax Latest: 2.0%
    • GPT 6 Luna: 2.5%Top

  • SciCode

    Scientific computing in Python

    • MiniMax Latest: 47.1%
    • GPT 6 Luna: 50.9%Top

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • MiniMax Latest: 42.4%
    • GPT 6 Luna: —

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • MiniMax Latest: 88.9%
    • GPT 6 Luna: —

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • MiniMax Latest: 83.0%Top
    • GPT 6 Luna: 78.3%

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

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Throughput
170.7 tok/s
Faster than 80% of 251 models
Time to first token
1.2 s
Quicker than 92% of 251 models
Prompt cache hits
80.3%
About $0.0854 per 1M input tokens with caching
Request success (24h)
99.9%
Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 87.5%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

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/minimax-latest",
    "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/minimax-latest",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

MiniMax

Other MiniMax models, starting with the same family.