MiniMax-M2

MiniMax-M2 is a compact, high-efficiency large language model optimized for end-to-end coding and agentic workflows. With 10 billion activated parameters (230 billion total), it delivers near-frontier intelligence across general reasoning, tool use, and multi-step task execution while maintaining low latency and deployment efficiency.

The model excels in code generation, multi-file editing, compile-run-fix loops, and test-validated repair, showing strong results on SWE-Bench Verified, Multi-SWE-Bench, and Terminal-Bench. It also performs competitively in agentic evaluations such as BrowseComp and GAIA, effectively handling long-horizon planning, retrieval, and recovery from execution errors.

Benchmarked by Artificial Analysis, MiniMax-M2 ranks among the top open-source models for composite intelligence, spanning mathematics, science, and instruction-following. Its small activation footprint enables fast inference, high concurrency, and improved unit economics, making it well-suited for large-scale agents, developer assistants, and reasoning-driven applications that require responsiveness and cost efficiency.

To avoid degrading this model's performance, MiniMax highly recommends preserving reasoning between turns. Learn more about using reasoning_details to pass back reasoning in our docs.

minimax/MiniMax-M2

Organization: MiniMax (minimax.io)
Family: minimax
Providers: 12
Context: 204,800
Output limit: 131,072
Knowledge: —
Release: 2025-10-27
Updated: 2025-10-27
Weights: Open
Input: text
Output: text
Capabilities: Tools, Reasoning, Temperature

Providers

Compare every cataloged offer for this model, including its provider, route, token limits, price, and capabilities.

14 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
Amazon Bedrock	minimax.minimax-m2	204,608	128,000	$0.30 / $1.20	Yes	Yes	Yes	View
Cortecs	minimax-m2	400,000	196,000	$0.349 / $1.405	Yes	Yes	Yes	View
DevPass (LLM Gateway)	minimax-m2	196,608	131,072	$0.20 / $1.00	Yes	Yes	—	View
Eden AI	minimax/MiniMax-M2	204,800	131,072	$0.30 / $1.20	Yes	Yes	No	View
Hugging Face	MiniMaxAI/MiniMax-M2	204,800	131,072	$0.30 / $1.20	Yes	Yes	—	View
Kilo Gateway	minimax/minimax-m2	196,608	176,947	$0.30 / $1.20	Yes	Yes	Yes	View
LLM Gateway	minimax/minimax-m2	196,608	131,072	$0.20 / $1.00	Yes	Yes	No	View
Merge Gateway	minimax/minimax-m2	204,800	8,192	$0.30 / $1.20	Yes	Yes	Yes	View
NanoGPT	minimax/minimax-m2	200,000	131,072	$0.17 / $1.53	Yes	No	No	View
Ofox	minimax/minimax-m2	204,800	131,072	$0.30 / $1.20	Yes	Yes	—	View
Googlevia OpenRouter	minimax/minimax-m2 unknown	196,608	176,947	$0.30 / $1.20	Yes	Yes	Yes	View
Minimaxvia OpenRouter	minimax/minimax-m2 fp8	204,800	131,072	$0.255 / $1.02	Yes	Yes	No	View
Novitavia OpenRouter	minimax/minimax-m2 fp8	204,800	131,072	$0.30 / $1.20	Yes	Yes	No	View
Vercel AI Gateway	minimax/minimax-m2	205,000	196,608	$0.30 / $1.20	Yes	Yes	—	View

Prices and limits apply to each offer. A dash means the value is not provided. OpenRouter offers show the hosting provider and routing channel separately.

Pricing

Compare token rates and cache charges across available routes and hosting configurations.

USD per million tokens

Input from$0.17Per 1M input tokens

Output from$1.00Per 1M output tokens

Available offers14Across 12 providers

Lowest listed input and output prices across the offers above; they may belong to different providers. These are base rates, before conditional pricing, discounts or additional fees.

Endpoint price comparisonLatest observed snapshot · 3 lowest-priced configurations

InputOutput

Minimaxminimax/fp8

$0.255$1.02

Googlegoogle-vertex

$0.30$1.20

Novitanovita/fp8

$0.30$1.20

minimax/minimax-m2Observed Oct 4, 2026, 11:31 AM UTC

Input: $0.30
Output: $1.20

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
Googlegoogle-vertex	$0.30	$1.20	—	—
Minimaxminimax/fp8	$0.255	$1.02	—	—
Novitanovita/fp8	$0.30	$1.20	$0.03	—

Each row is a hosting configuration. Open an offer in Providers for its full pricing conditions and additional fees. Endpoint observations may differ from the routing catalog's base price.

Performance

Review reported latency and output-throughput percentiles from the latest observed rolling window.

Latest 30-minute observation

Time to first token · milliseconds

Latency distributionRolling 30-minute snapshot · 3 endpoints

P50P75P90P99

Googlegoogle-vertex

692–3,620.91 ms

Minimaxminimax/fp8

903–1,744.3 ms

Novitanovita/fp8

1,430–7,863.15 ms

All latency percentiles 3 endpoints

Endpoint	P50	P75	P90	P99
Googlegoogle-vertex	692 ms	1,227.5 ms	1,915.8 ms	3,620.91 ms
Minimaxminimax/fp8	903 ms	1,420.5 ms	1,609.4 ms	1,744.3 ms
Novitanovita/fp8	1,430 ms	1,821.5 ms	4,346 ms	7,863.15 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 3 endpoints

P50P75P90P99

Googlegoogle-vertex

118–170.2 t/s

Minimaxminimax/fp8

67.5–113.32 t/s

Novitanovita/fp8

69–103.7 t/s

All throughput percentiles 3 endpoints

Endpoint	P50	P75	P90	P99
Googlegoogle-vertex	118 t/s	122 t/s	148 t/s	170.2 t/s
Minimaxminimax/fp8	67.5 t/s	79.75 t/s	98.2 t/s	113.32 t/s
Novitanovita/fp8	69 t/s	78 t/s	83 t/s	103.7 t/s

Percentiles describe the observed request distribution. These are snapshots of a rolling window, not a historical time series.

Uptime

Compare successful request rates across the latest 5-minute, 30-minute, and 24-hour observations.

Successful requests by window

Availability by endpointCurrent overlapping windows · 3 endpoints

Endpoint5 minutes30 minutes24 hours

Googlegoogle-vertexObserved Oct 4, 2026, 11:31 AM UTC

—

97.38%

Minimaxminimax/fp8Observed Oct 4, 2026, 11:31 AM UTC

—

99.89%

Novitanovita/fp8Observed Oct 4, 2026, 11:31 AM UTC

—

99.75%

Reported by OpenRouter at the time of collection; rate-limited requests are excluded. A dash means no measurement was reported. This is not a live availability check.

Benchmarks

Review attributed evaluation results on each source's original scale.

External evaluations

minimax/minimax-m2Observed Oct 4, 2026, 11:31 AM UTC

Design Arena	Category	Elo	Win rate	Rank
models	3d	1,113	48.3%	80
models	codecategories	1,141	48.1%	84
models	dataviz	1,149	50%	76
models	gamedev	1,131	48.1%	80
models	svg	1,099	55.3%	59
models	uicomponent	1,143	49.2%	77
models	website	1,149	48%	83

Sources: Artificial Analysis and Design Arena, via OpenRouter. Scores retain their original scales and are not OpenAI Suite ratings.

Detailed evaluations

Evaluation	Configuration	Source	Results	Details
models · gamedev	MiniMax M2 Stableminimax/minimax-m2	Design Arena	1,131 Elo	Details
models · uicomponent	MiniMax M2 Stableminimax/minimax-m2	Design Arena	1,143 Elo	Details
models · codecategories	MiniMax M2 Stableminimax/minimax-m2	Design Arena	1,141 Elo	Details
models · svg	MiniMax M2 Stableminimax/minimax-m2	Design Arena	1,099 Elo	Details
models · dataviz	MiniMax M2 Stableminimax/minimax-m2	Design Arena	1,149 Elo	Details
models · website	MiniMax M2 Stableminimax/minimax-m2	Design Arena	1,149 Elo	Details
models · 3d	MiniMax M2 Stableminimax/minimax-m2	Design Arena	1,113 Elo	Details

Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Oct 4, 2026.

Technical details & API

Explore specifications, supported parameters, reasoning controls, and API identifiers for each OpenRouter route.

Current route specifications

minimax/minimax-m2Observed Oct 4, 2026, 11:31 AM UTC

Context: 204,800
Maximum output: 176,947
Tokenizer: Other
Input: text
Output: text
Moderation: No
Added to OpenRouter: 2025-10-23

Base URLhttps://openrouter.ai/api/v1

Model IDminimax/minimax-m2

Version IDminimax/minimax-m2

Weights IDMiniMaxAI/MiniMax-M2

Reasoning

Required: Yes
Enabled by default: Not reported
Default effort: —
Supported efforts: —

Supported parameters

frequency_penalty
include_reasoning
logprobs
max_tokens
presence_penalty
reasoning
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_logprobs
top_p

Default parameters

top p: 0.95
temperature: 1

Frequently asked questions

Answers based on the model metadata and serving offers currently in the catalog.

How much does MiniMax-M2 cost?

Listed rates start at $0.17 per million input tokens and $1.00 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer MiniMax-M2?

There are 14 cataloged offers across 12 providers. Compare model IDs, prices and limits in Providers.

What is the context window?

The cataloged model context is 204,800 tokens. Each serving endpoint may apply a different limit.

Does it support tools and structured output?

Tool calling: Yes. Structured output: Not reported. Support can vary by endpoint.

Resources

Hugging Face ↗

Catalog history · 2 recorded revisions

Metadata observations since this model was first cataloged. These are not model release versions.

Sep 14, 2026 · 1ef7ea63ab84
Sep 12, 2026 · 91ff1193c7b8