MiniMax M2
3611°
Run this model
Offers / Ads
MiniMax | 204,800 token context | $0.26/M input, $1.02/M output. 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…MiniMax | 204,800 token context | $0.26/M input, $1.02/M output. 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 AnalysisOpens in new tab, 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 docsOpens in new tab.
Guides for this model
Best local LLM by use case (October 2026): the right open model for every job
Open weights vs open source vs API-only: what you are actually allowed to do
How much VRAM do I actually need for local AI?
What is a GGUF file? Quantized models explained (Q4, Q8, fp8, NF4)
Captain's notes
Notes from 2026-10-13
MiniMax M2 is a mixture-of-experts model from MiniMax with 230 billion total parameters but only 10 billion active at inference, giving it the throughput of a small model with reasoning closer to frontier scale. It is built specifically for coding and agentic work: multi-file edits, compile-run-fix loops, long-horizon tool chains, and browser-based retrieval. Open weights under MIT and priced at $0.30 per million input tokens, it suits developers who need capable agents without frontier-tier bills.
Known for
- Strong SWE-Bench and Terminal-Bench scores among open-source models
- Long-horizon agentic planning across shell, browser, and code runners
- MIT licence: open weights, commercial use allowed
- 204,800-token context window for large codebases and multi-turn agents
- Low activation footprint keeps latency and cost down at scale
Watch out for
- MiniMax recommends passing reasoning back between turns; skipping this degrades performance noticeably.
- 230 billion total parameters makes self-hosting hardware-intensive despite the small active footprint.
- API pricing is published; free-tier availability is described as limited-time only.
Prompts to try
Multi-file refactor task
You have a Python web API split across three files: routes.py, models.py, and db.py. The database layer uses raw SQL strings scattered through all three files. Refactor the code to centralise all queries into a repository class in db.py, update the other files accordingly, and write a pytest test for each repository method. Show the full updated files.
Compile-run-fix loop
Here is a Rust function that is supposed to parse a comma-separated list of integers and return their sum, but it fails to compile. Identify every error, explain why each one occurs, produce a corrected version that compiles cleanly, and add a unit test that covers empty input, a single value, and a list with whitespace around the commas.
fn sum_csv(input: &str) -> i32 {
input.split(',').map(|s| s.parse()).sum()
}
Long-horizon agent plan
You are an agent with access to a shell, a web browser, and a code runner. A user wants to know which of the top five most-downloaded Python packages on PyPI last month has the highest ratio of open GitHub issues to closed ones. Write the step-by-step plan you would execute, the exact commands or requests at each step, how you would handle a failed HTTP request, and what final output you would return.
Compare with
- MiniMax M1API-only model1,000,000 token context · $0.55/M input, $2.20/M output
- MiniMax M3Open-weights model1,048,576 token context · $0.30/M input, $1.20/M output
- MiniMax M2.5Open-weights model204,800 token context · $0.27/M input, $0.95/M output
- MiniMax M2.1Open-weights model204,800 token context · $0.30/M input, $1.20/M output
More text models from minimax
Questions people ask
- Are the weights open and can I run MiniMax M2 locally?
- Yes. MiniMax M2 is released under the MIT licence and the weights are on Hugging Face at MiniMaxAI/MiniMax-M2. Bear in mind that the full model has 230 billion parameters, so self-hosting requires substantial hardware even though only 10 billion parameters are active per forward pass.
- What does MiniMax M2 cost and how long is its context?
- Via the API it is $0.30 per million input tokens and $1.20 per million output tokens, with a 204,800-token context window.
- How does MiniMax M2 compare to other open coding models?
- Its SWE-Bench Verified score of 69.4 and Terminal-Bench score of 46.3 place it ahead of DeepSeek-V3.2 and Kimi K2 on those tasks, while its BrowseComp score of 44 is notably higher than most open alternatives, though closed models like Claude Sonnet 4.5 lead on several of the same benchmarks.
Captain's verdictA rare open-weights model that competes on real coding and agent benchmarks without demanding frontier-tier infrastructure or pricing.
Model card on Hugging Face
Captain writes these notes from the model's own description and model card. Prices and context sizes are the provider's at the time of writing and can change.
Ranking badge
Updated Oct 22, 2026
Made this model? Show this badge on your page.
Paste it into your Hugging Face card or Civitai description.
How this score adds up
This model has no popularity data from any source yet, so it carries a stand-in number below 7,000° that keeps the lists complete. It never ranks above a model with data. The breakdown appears here as soon as a source reports it.
Glossary: the terms used here
- Heat score
- One number for how much a model is used and wanted right now, from 0° to 9,999°. Every list on this site that says hottest is sorted by it and nothing else.
- Floor
- The 7,000° every model with real data starts from. The scale runs from 0° to 9,999°; the part below the floor is reserved for stand-ins (models without data), so that a model with data always ranks above one without. The floor is the same for everyone and decides nothing; the degrees above it are the measurement.
- Download-equivalents
- Everything the score treats as a download, added up: downloads themselves, each torrent seeder as 1,000, each AI Horde image and each run on a generation site as one.
- Log scale
- Counting by tens, not by ones: going from 1,000 to 10,000 downloads adds the same degrees as going from 10,000 to 100,000. It keeps giants from crowding out everything else.
- Likes
- Thumbs up and likes from every source, worth half a download each on the log scale.
- Adoption rate
- How many downloads a day a model gets right now, from the difference between two readings of a source. A quarter of the weight of the download term.
- Measured and estimated
- Measured means two readings of a source exist and the difference is real. Estimated means only today's total is known and the past is spread over the model's life; the site says which one it is showing.
- Seeder
- Someone who keeps a copy of the weights online for others to download over a torrent, or a hosted mirror doing the same. Each one counts as 1,000 downloads.
- AI Horde
- A volunteer network that generates images for free on donated computers. The number of images it has made with a model is public, and each counts as one download.
- Generation sites
- Sites that host their own copies of a model and let people generate with it, such as PixAI, SeaArt and Yodayo. Their downloads count like Civitai's and each generation run as one download.
- OpenRouter tokens
- A text model served through an API, such as Claude or GPT, has no weights to download, so its usage is measured in tokens: the pieces of text it read and wrote through OpenRouter, a gateway many apps use, over the last 30 days. Every 10,000,000 tokens count as one download. Source: OpenRouter (openrouter.ai/rankings), CC BY 4.0.
- Full model (checkpoint)
- The weights themselves, the file that generates on its own. It gets a bonus because add-ons depend on it.
- LoRA and embedding
- Small add-on files that steer someone else's weights toward a style, a character or a concept. They cannot run alone.
- Family and weights
- The base a model is built on, such as SD 1.5, SDXL, Flux or Wan. Every model inherits the release date of its family's weights.
- Age
- Degrees taken off for how long ago the family's weights came out: 150° a year, up to 750°. It keeps old downloads that had years to pile up from freezing the top of the list.
- Family boost
- A small, bounded nudge for LoRAs of a family this site is pushing forward right now, always listed in the breakdown.
- Stand-in score
- A stable number below the floor given to a model no source reports on yet, so lists stay complete. It never ranks above a model with data.
- New! (grace period)
- A model less than 7 days old shows New! instead of a score and holds no place in the rankings. Its score is computed from day one, but a model one day old grows faster than anything, so without the wait the week's #1 would always be whatever came out yesterday rather than what held up for a week.
- Trend
- How a model's growth over a period compares with the typical model's growth over the same period. Faster than typical reads as a rise, slower as a fall.
- Typical model
- The middle of the catalog: half the models with a figure grew faster over the period, half slower.
MiniMax M2 is a text model for the Text / LLM family listed on ModelTrends.ai, a read-only catalog of open source AI image, video and text models. Compare it with other Text / LLM models and check its heat score to see how it is trending.
Browse by Model Family
Text / LLM Models
434
Anima Models
528
Chroma Models
19
Fibo Models
2
Flux Models
1,496
Flux 2 / Klein Models
1,239
Flux 3 Models
3
Moss Models
3
Iris Models
1
MiniMax H3 Models
364
Hunyuan Models
204
Ideogram Models
20
Krea2 Models
717
Qwen Models
20
Qwen2 Models
47
Ltx2 Models
153