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, created on February 12, 2026, demonstrates a strong overall performance profile, particularly excelling in reliability. It exhibits exceptional reliability with a 99% success rate across benchmarks, indicating minimal technical failures. In terms of speed, M2.5 shows moderate performance, ranking in the 31st percentile, while its pricing is also moderate, placing it in the 35th percentile. The model showcases significant strengths in several key areas. It achieves perfect accuracy in both General Knowledge and Ethics benchmarks, with the latter also being the most accurate and fastest among models at its price point. M2.5 also performs very well in Coding (93.0% accuracy, 82nd percentile) and Reasoning (94.0% accuracy, 80th percentile), aligning with its description as a SOTA large language model for real-world productivity and complex digital working environments. Its 204800 context length further supports its ability to handle extensive tasks. A notable weakness is observed in its Hallucinations benchmark, where it scores 88.0% accuracy, placing it in the 35th percentile. This suggests room for improvement in appropriately acknowledging uncertainty. Instruction Following and Mathematics benchmarks show solid, though not top-tier, performance at 65.0% and 91.0% accuracy respectively. Overall, M2.5 is a highly reliable and capable model, particularly strong in knowledge, ethics, coding, and reasoning, making it well-suited for diverse professional applications.
Model Pricing
Current Pricing
| Feature | Price (per 1M tokens) |
|---|---|
| Prompt | $0.19 |
| Completion | $1.15 |
| Input Cache Read | $0.095 |
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 | $0.19 / 1M tokens | $1.15 / 1M tokens |
|
Minimax
|
Minimax | minimax/minimax-m2.5-20260211 | 204K | $0.19 / 1M tokens | $1.15 / 1M tokens |
|
Minimax
|
Minimax | minimax/minimax-m2.5-20260211 | 204K | $0.6 / 1M tokens | $2.4 / 1M tokens |
|
SiliconFlow
|
SiliconFlow | minimax/minimax-m2.5-20260211 | 196K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
Fireworks
|
Fireworks | minimax/minimax-m2.5-20260211 | 196K | $0.19 / 1M tokens | $1.15 / 1M tokens |
|
AtlasCloud
|
AtlasCloud | minimax/minimax-m2.5-20260211 | 196K | $0.295 / 1M tokens | $1.2 / 1M tokens |
|
Parasail
|
Parasail | minimax/minimax-m2.5-20260211 | 196K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
Chutes
|
Chutes | minimax/minimax-m2.5-20260211 | 204K | $0.19 / 1M tokens | $1.15 / 1M tokens |
|
Inceptron
|
Inceptron | minimax/minimax-m2.5-20260211 | 196K | $0.28 / 1M tokens | $1.1 / 1M tokens |
|
Together
|
Together | minimax/minimax-m2.5-20260211 | 196K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
NextBit
|
NextBit | minimax/minimax-m2.5-20260211 | 196K | $0.2 / 1M tokens | $1.2 / 1M tokens |
|
Chutes
|
Chutes | minimax/minimax-m2.5-20260211 | 196K | $0.19 / 1M tokens | $1.15 / 1M tokens |
|
Novita
|
Novita | minimax/minimax-m2.5-20260211 | 204K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
SambaNova
|
SambaNova | minimax/minimax-m2.5-20260211 | 163K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
Friendli
|
Friendli | minimax/minimax-m2.5-20260211 | 196K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
Clarifai
|
Clarifai | minimax/minimax-m2.5-20260211 | 196K | $0.19 / 1M tokens | $1.15 / 1M tokens |
|
Venice
|
Venice | minimax/minimax-m2.5-20260211 | 198K | $0.34 / 1M tokens | $1.19 / 1M tokens |
|
Ionstream
|
Ionstream | minimax/minimax-m2.5-20260211 | 196K | $0.2 / 1M tokens | $1.17 / 1M tokens |
|
DeepInfra
|
DeepInfra | minimax/minimax-m2.5-20260211 | 196K | $0.27 / 1M tokens | $0.95 / 1M tokens |
|
Nebius
|
Nebius | minimax/minimax-m2.5-20260211 | 196K | $0.3 / 1M tokens | $1.2 / 1M tokens |
|
AkashML
|
AkashML | minimax/minimax-m2.5-20260211 | 196K | $0.3 / 1M tokens | $1.18 / 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 M2.7 Unavailable | Mar 18, 2026 | — | 204K |
Text input
Text output
|
★ | ★★★★ | $$$$ |
| MiniMax: MiniMax M2.7 | Mar 18, 2026 | — | 204K |
Text input
Text output
|
★ | ★★★★ | $$$$ |
| MiniMax: MiniMax M2-her | Jan 23, 2026 | — | 65K |
Text input
Text output
|
★★ | ★ | $$$ |
| MiniMax: MiniMax M2.1 | Dec 22, 2025 | ~10B | 204K |
Text input
Text output
|
★ | ★★★★ | $$$$$ |
| MiniMax: MiniMax M2 | Oct 23, 2025 | ~230B | 196K |
Text input
Text output
|
★ | ★★★ | $$$$$ |
| MiniMax: MiniMax M1 | Jun 17, 2025 | — | 1M |
Text input
Text output
|
★ | ★★★★ | $$$$$ |
| MiniMax: MiniMax M1 (extended) Unavailable | Jun 17, 2025 | — | 128K |
Text input
Text output
|
★ | ★ | $$$$ |
| MiniMax: MiniMax-01 | Jan 14, 2025 | ~456B | 1M |
Text input
Image input
Text output
|
★★★ | ★★ | $$$ |