Requesty
MiniMax

minimax-m3

MiniMax M3 is a frontier multimodal model with a 1M-token context window built on MiniMax Sparse Attention (MSA). It reaches frontier-level performance on coding and agentic tasks, surpassing GPT-5.5 and Gemini 3.1 Pro on SWE-Bench Pro and approaching Claude Opus 4.7. It natively supports image and video input and is the first open-weight model to combine frontier coding, ultra-long context, and native multimodality.

VisionReasoningTool callingCachingJSON schema
Try minimax-m3CompareDocs

Specifications

Context window1M tokens
Max output128K tokens
API typechat
AddedJun 1, 2026
Model IDminimaxi/minimax-m3
Data retentionYes
Used for trainingUnknown
Provider locationπŸ‡ΈπŸ‡¬ Singapore
Privacy policyMiniMax Privacy Policy→

Benchmarks

Released 2026-06-01
Coding Indexcoding
43.4%

Artificial Analysis Coding Index β€” a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.

GPQA Diamondreasoning
92.9%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

Intelligence Indexreasoning
54.7%

Artificial Analysis Intelligence Index β€” a composite of multiple evaluations measuring overall model capability.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality β€” always test on your own workload.

Pricing

Input / 1M
$0.30
Output / 1M
$1.20
Cache write
β€”
Cache read / 1M
$0.06
Estimated cost
100K input + 10K output$0.0420
1M input + 100K output$0.42
10M input + 1M output$4.20

Requesty charges exactly what the upstream provider charges β€” no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to minimaxi/minimax-m3.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="minimaxi/minimax-m3",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other MiniMax models

MiniMax-M2.7
200K ctx$0.30/1M
Coding 42
MiniMax-M2.7-highspeed
200K ctx$0.60/1M
MiniMax-M2.5-highspeed
200K ctx$0.60/1M
MiniMax-M2.5
200K ctx$0.30/1M
Coding 37
MiniMax-M2
200K ctx$0.30/1M
Coding 29

Frequently asked questions

How much does minimax-m3 cost?
minimax-m3 is priced at $0.30 per million input tokens and $1.20 per million output tokens when accessed via Requesty. Prompt caching is supported, which can cut effective input cost by up to 90% on repeated context. Requesty charges exactly what the upstream provider charges β€” we don't add markup.
What is the context window of minimax-m3?
minimax-m3 has a context window of 1M tokens, with a maximum output of 128K tokens per response. That's roughly 1,333 words of input you can fit in a single prompt.
How does minimax-m3 perform on benchmarks?
minimax-m3 scores 92.9% on GPQA Diamond, 88.9% on τ²-Bench, 54.7% on Intelligence Index. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can minimax-m3 do?
minimax-m3 supports vision input, tool calling, extended reasoning, prompt caching, structured outputs (JSON schema). You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use minimax-m3 with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "minimaxi/minimax-m3". The Quickstart above shows Python, JavaScript and cURL snippets.

Access minimax-m3 through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

Get started freeAll MiniMax models