MiniMax M2.5
6216°
Run this model
Offers / Ads
MiniMax | 204,800 token context | $0.27/M input, $0.95/M output. MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diverse range of complex real-world digital working environments, M2.5 builds upon the coding expertise of M2.1 to extend into general office work, reaching fluency in generating and operating Word, Excel, and Powerpoint files, context switching between diverse software environments, and working across different agent and human teams. Scoring 80.2% on SWE-Bench Verified, 51.3% on Multi-SWE-Bench, and 76.3% on BrowseComp, M2.5 is also more token efficient than previous generations, having been trained to optimize its actions and output through planning.
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 M2.5 is a frontier-class open-weights model from MiniMax, built specifically for agentic work: coding across the full development lifecycle, browser-based search, and real office-file tasks in Word, Excel, and PowerPoint. Its reinforcement-learning training in real-world environments gives it unusually strong tool-use and planning discipline. At $0.27/M input and $1.08/M output with a 204,800-token context, it undercuts most rivals at comparable benchmark levels.
Known for
- 80.2% on SWE-Bench Verified, competitive with top closed models on multiple harnesses.
- Plans code like an architect before writing a line, decomposing structure and features first.
- Trained on 13+ languages across web, Android, iOS, and Windows full-stack projects.
- Generates and operates Word, Excel, and PowerPoint files natively, not just text about them.
- 76.3% on BrowseComp; strong autonomous web search and tool-calling in agentic pipelines.
Watch out for
- Open weights under a Modified-MIT licence — check the exact terms before commercial deployment.
- Frontier-scale model; local inference will demand serious hardware. VRAM requirements are not published.
- Context tops out at 204,800 tokens — long but not in the 1M-plus tier some rivals now offer.
Prompts to try
Full-stack feature from scratch
You are building a task-management web app. Design the architecture first: list the components, data models, and API endpoints. Then implement a REST endpoint in Node.js with Express that creates a new task, validates the input, and stores it in a SQLite database. Include error handling and a brief test using curl.
Multi-step agentic plan
I need to research the top five open-source vector databases by GitHub stars, compare their licensing, and produce a recommendation for a startup that wants to self-host. Break the task into steps, state what tools or searches you would run at each step, then execute the plan and give me a final comparison table.
Excel logic in plain code
Write a Python script using openpyxl that opens a file called sales.xlsx, reads a sheet named Q1, sums the values in column C for every row where column B equals 'EMEA', and writes the result into cell F2 with the label 'EMEA Total' in E2. Handle the case where the sheet does not exist.
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 M2.1Open-weights model204,800 token context · $0.30/M input, $1.20/M output
- MiniMax M2Open-weights model204,800 token context · $0.26/M input, $1.02/M output
More text models from minimax
Questions people ask
- Are the weights open and can I run M2.5 locally?
- Yes, the weights are available on Hugging Face under a Modified-MIT licence, so local deployment is possible — though MiniMax has not published the parameter count or minimum VRAM requirements, so check the repository before committing to hardware.
- What does M2.5 cost and what is the context window?
- Via the API it is $0.27 per million input tokens and $1.08 per million output tokens, with a 204,800-token context window.
- How does M2.5 differ from M2.1?
- M2.5 extends M2.1's coding focus into general office work — native Word, Excel, and PowerPoint file handling — and is reported to complete agentic tasks 37% faster than M2.1 while scoring higher on SWE-Bench Verified and BrowseComp.
Captain's verdictA serious agentic coding model at a price that makes continuous autonomous runs genuinely affordable — the office-file fluency is a rare differentiator.
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 M2.5 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