MiniMax M3

minimax/minimax-m3
Featured
VisionToolsJSONReasoning
by MiniMax · 2026-05-31

MiniMax-M3 is MiniMax's flagship open-weight foundation model and the first to combine three frontier capabilities at once: frontier-level coding and agentic performance, a 1M-token context window, and native multimodality. It accepts text, image, and video inputs with text output, and is powered by the proprietary MiniMax Sparse Attention (MSA) architecture, which sustains up to 1M tokens of context (with a guaranteed minimum of 512K) - the foundation for long-range agent tasks, long-horizon coding, and long-video understanding. Multimodality is a native core capability rather than an add-on: the data pipeline was rebuilt to scale pretraining to 100T+ tokens with multimodal training from step zero, deeply aligning textual and visual semantic spaces. M3 achieves top-tier results across coding and agentic benchmarks spanning software engineering, terminal execution, and autonomous browsing (scoring 83.5 on BrowseComp), with autonomous task decomposition, tool invocation, and multi-step reasoning. It is well suited to AI coding assistants, automated workflows, and long-running asynchronous agent pipelines where coherence over extended sessions matters.

ctx1.05M tokens
Max output512K
Inputtext + image + video
Outputtext
p50 TTFT1.78 s
INPUT$0.30/ 1M tokens
OUTPUT$1.20/ 1M tokens
p50 TTFT1.78 s7d
p95 TTFT9.07 s7d
TRAFFIC42.2Mtokens / 7d

MiniMax M3 is a large language model developed by Minimax, accessible through OrcaRouter's OpenAI-compatible API. It accepts text, image, and video inputs, and has a context window of 1,048,576…

What is MiniMax M3 and who is it for?

What input modalities does MiniMax M3 support?

What is the context window and output limit of MiniMax M3?

Code samples

Call from any SDK

OpenAI-compatible — keep the SDK you already use

  • OpenAI SDKhttps://api.orcarouter.ai/v1
from openai import OpenAI

client = OpenAI(
    base_url="https://api.orcarouter.ai/v1",
    api_key="$ORCAROUTER_API_KEY",
)

response = client.chat.completions.create(
    model="minimax/minimax-m3",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Supported parameters

  • include_reasoning
  • max_tokens
  • reasoning
  • response_format
  • temperature
  • tool_choice
  • tools
  • top_p

Pricing

Input / 1M tokens$0.300
Output / 1M tokens$1.20
Cache read / 1M$0.060
CurrencyUSD

Cost calculator

Tokens / month10MM
Input share70%%
Estimated / month $5.70 · With prompt caching $4.86

Estimate based on list price

Token & cost estimator

Input tokens: 20Cost per request: $0.000606

Estimate only — actual token counts depend on the provider's tokenizer.

Performance

p50 TTFT
1.78 s
Output speed
105 tok/s
p95 TTFT
9.07 s
Error rate
0.16%

Public benchmarks

56.2
AA Coding
Better than 80% of models compared
#21 of 106
62.2
AA Intelligence
Better than 88% of models compared
#13 of 110
59.2
AA Math
Better than 41% of models compared
#48 of 81
BankerToolBench
76.1
BrowseComp
83.5
GDPval rubrics
76.7
GPQA Diamond
55.2 index
KernelBench Hard
25.8
MCP Atlas
74.2
MMLU-Pro
68.2 index
OSWorld-verified
70.0
SVG-Bench
63.7
SWE Bench Pro
59.0
Terminal Bench 2.1
66.0
VIBE V2
50.7
τ²-Bench
54.2 index
Source: artificialanalysis.ai, minimax.io

How it compares

MiniMax M3MiniMax M2.7MiniMax M2.7 highspeedMiniMax M2.5
Input $/M$0.30$0.30$0.60$0.30
Output $/M$1.20$1.20$2.40$1.20
Context1.0M205K205K205K
Quality9/108/108/107/10
Compare side-by-sideCompare side-by-sideCompare side-by-sideCompare side-by-side

FAQ

What is the cost per token for MiniMax M3?
MiniMax M3 costs $0.30 per 1 million input tokens and $1.20 per 1 million output tokens. These rates are billed at the provider rate with zero markup from OrcaRouter.
What is the context window size of MiniMax M3?
MiniMax M3 has a context window of 1,048,576 tokens (approximately 1 million tokens). It also has a maximum output limit of 512,000 tokens.
What are the main strengths of MiniMax M3?
Its main strengths are a very large context window (1,048,576 tokens), multimodal input support (text, image, video), and a strong score of 83.5 on the BrowseComp benchmark for web-based information retrieval.
How does MiniMax M3 compare to models with smaller context windows?
MiniMax M3's context window is significantly larger than most models (e.g., 128K or 200K tokens), enabling processing of entire documents or videos in a single request. This can reduce complexity and cost for long-context tasks, though its per-token pricing may be higher than smaller models on a per-token basis.
How is data handled when using MiniMax M3 through OrcaRouter?
Data is processed via OrcaRouter's infrastructure. For details on data retention, encryption, and privacy, consult OrcaRouter's privacy policy and Minimax's data handling terms. No specific data handling features are provided in the available facts.
How do I call MiniMax M3 via an OpenAI-compatible API?
Send requests to base URL https://api.orcarouter.ai/v1 with model ID "minimax/minimax-m3". Use your OrcaRouter API key in the Authorization header. The API is fully compatible with the OpenAI SDK; you simply change the base URL and model name.
What input formats does MiniMax M3 support?
MiniMax M3 supports text, image, and video as input modalities. Images and videos can be provided as URLs or base64-encoded data within the content array, following OpenAI's multimodal message format.
Is there a limit on the output length?
Yes, the maximum output length is 512,000 tokens. If you set a `max_tokens` parameter in your API call, it must not exceed this value.
Does OrcaRouter add any markup to the model's pricing?
No. The pricing of $0.30 per 1M input tokens and $1.20 per 1M output tokens is the provider rate with zero markup from OrcaRouter.
What is the BrowseComp score and what does it mean?
MiniMax M3 scored 83.5 on BrowseComp, a benchmark that tests a model's ability to navigate web pages and extract information. This indicates strong performance in web-based question answering.

Embed this badge

MiniMax: MiniMax M3$0.30/M in1777ms p50via OrcaRouter
HTML <a href="https://www.orcarouter.ai/models/minimax/minimax-m3" target="_blank"> <img src="https://www.orcarouter.ai/embed/minimax/minimax-m3.svg" alt="MiniMax: MiniMax M3 on OrcaRouter" /> </a>
Markdown [![MiniMax: MiniMax M3](https://www.orcarouter.ai/embed/minimax/minimax-m3.svg)](https://www.orcarouter.ai/models/minimax/minimax-m3)

Model card as data

GET /api/public/models/minimax/minimax-m3Open
Machine-readable:/llms.txt/llms-full.txt