Author's Description
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.
Key Specifications
Supported Parameters
This model supports the following parameters:
Features
This model supports the following features:
Performance Summary
MiniMax M2.5 demonstrates moderate speed performance, ranking in the 30th percentile across benchmarks, and offers moderate pricing, placing it in the 32nd percentile. A standout feature is its exceptional reliability, achieving a 99% success rate, indicating minimal technical failures. The model exhibits strong performance in several key areas. It achieves perfect accuracy in both General Knowledge and Ethics, with the added distinction of being the most accurate model at its price point and among models of similar speed in these categories. M2.5 also shows robust capabilities in Coding (93.0% accuracy, 83rd percentile) and Reasoning (94.0% accuracy, 84th percentile), aligning with its description as a SOTA LLM for productivity and complex real-world tasks. Its Instruction Following is solid at 65.0% accuracy (68th percentile), and Email Classification is strong at 98.0% accuracy (61st percentile). A notable weakness is its Hallucinations benchmark, where it scores 88.0% accuracy (36th percentile), suggesting room for improvement in acknowledging uncertainty. While its Mathematics performance is respectable at 91.0% accuracy (64th percentile), it is not a top-tier result. Overall, M2.5 is a highly reliable model with significant strengths in knowledge, ethics, coding, and reasoning, making it well-suited for diverse professional applications despite moderate speed and a slight tendency towards hallucination.
Model Pricing
Current Pricing
| Feature | Price (per 1M tokens) |
|---|---|
| Prompt | $0.3 |
| Completion | $1.2 |
| Input Cache Read | $0.03 |
Price History
Available Endpoints
| Provider | Endpoint Name | Context Length | Pricing (Input) | Pricing (Output) |
|---|---|---|---|---|
|
Minimax
|
Minimax | minimax/minimax-m2.5-20260211 | 204K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
Novita
|
Novita | minimax/minimax-m2.5-20260211 | 204K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
Minimax
|
Minimax | minimax/minimax-m2.5-20260211 | 204K | $0.3 / 1M tokens | $2.4 / 1M tokens |
Benchmark Results
| Benchmark | Category | Reasoning | Strategy | Free | Executions | Accuracy | Cost | Duration |
|---|
Other Models by minimax
|
|
Released | Params | Context |
|
Speed | Ability | Cost |
|---|---|---|---|---|---|---|---|
| MiniMax: MiniMax M2-her | Jan 23, 2026 | — | 65K |
Text input
Text output
|
★★ | ★ | $$$ |
| MiniMax: MiniMax M2.1 | Dec 22, 2025 | ~10B | 204K |
Text input
Text output
|
★ | ★★★★ | $$$$$ |
| MiniMax: MiniMax M2 | Oct 23, 2025 | ~230B | 196K |
Text input
Text output
|
★ | ★★★ | $$$$$ |
| MiniMax: MiniMax M1 | Jun 17, 2025 | — | 1M |
Text input
Text output
|
★ | ★★★★ | $$$$$ |
| MiniMax: MiniMax M1 (extended) Unavailable | Jun 17, 2025 | — | 128K |
Text input
Text output
|
★ | ★ | $$$$ |
| MiniMax: MiniMax-01 | Jan 14, 2025 | ~456B | 1M |
Image input
Text input
Text output
|
★★★ | ★★ | $$$ |