MiniMax model product
MiniMax M2.5
5 is a 230B-parameter MoE large language model with 10B activation parameters, designed for agent scenarios and supporting programming, tool calling, search, and office productivity.
Updated Oct 23, 2026. Default version: MiniMax M2.5
LLMBoard Score53.3MiniMax M2.5
Coverage60%5 benchmark families
Context window1MTokens
Official input price$0.30MiniMax (minimax.io) API
MiniMax M2.5 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
MiniMax M2.5 Benchmark Results
Benchmark scores for MiniMax M2.5.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkBFCL_v3_MultiTurn | Score76.80% | Rank01 | Participants2 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkMEWC | Score74.40% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkMulti-SWE-Bench | Score51.30% | Rank02 | Participants7 | Percentile83.33% | EvidenceC | Evaluated |
| BenchmarkVIBE-Pro | Score54.20% | Rank02 | Participants2 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkGDPval-MM | Score59.00% | Rank03 | Participants3 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Verified | Score80.20% | Rank16 | Participants120 | Percentile87.40% | EvidenceC | Evaluated |
| BenchmarkBrowseComp | Score76.30% | Rank29 | Participants70 | Percentile59.42% | EvidenceC | Evaluated |
| BenchmarkAA IFBench | Score71.63% | Rank38 | Participants168 | Percentile77.84% | EvidenceB | EvaluatedN/A |
| BenchmarkSWE-Bench Pro | Score55.40% | Rank44 | Participants63 | Percentile30.65% | EvidenceC | Evaluated |
| BenchmarkAA GPQA Diamond | Score84.85% | Rank69 | Participants199 | Percentile65.66% | EvidenceB | EvaluatedN/A |
| BenchmarkAA LCR v1.1 | Score73.33% | Rank73 | Participants193 | Percentile62.50% | EvidenceB | EvaluatedN/A |
| BenchmarkLM Arena Webdev | Score1,386.34 rating | Rank76 | Participants107 | Percentile29.25% | EvidenceA | Evaluated |
| BenchmarkAA Omniscience Accuracy | Score26.18% | Rank79 | Participants201 | Percentile61.00% | EvidenceB | EvaluatedN/A |
| BenchmarkAA CritPt | Score1.14% | Rank86 | Participants200 | Percentile57.29% | EvidenceB | EvaluatedN/A |
| BenchmarkAA HLE Text No Tools | Score20.53% | Rank88 | Participants200 | Percentile56.28% | EvidenceB | EvaluatedN/A |
| BenchmarkLM Arena Text Factuality | Score1,385.35 rating | Rank119 | Participants131 | Percentile9.23% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text Style Control | Score1,390.75 rating | Rank128 | Participants218 | Percentile41.47% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text | Score1,359.22 rating | Rank139 | Participants218 | Percentile36.41% | EvidenceA | Evaluated |
MiniMax M2.5 Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext factuality | Categorypolish | Rank51 | Rating / score1,352.64 | Votes634 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev-html | Rank67 | Rating / score1,413.57 | Votes965 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryfrench | Rank69 | Rating / score1,399.21 | Votes923 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryspanish | Rank72 | Rating / score1,376.93 | Votes989 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev-react | Rank72 | Rating / score1,373.90 | Votes6,619 | ObservationsN/A | Result date |
| Arenawebdev | Categoryoverall | Rank76 | Rating / score1,386.34 | Votes7,621 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev | Rank76 | Rating / score1,386.34 | Votes7,621 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry mathematical | Rank85 | Rating / score1,387.75 | Votes1,692 | ObservationsN/A | Result date |
| Arenatext factuality | Categorychinese | Rank86 | Rating / score1,418.51 | Votes1,920 | ObservationsN/A | Result date |
| Arenatext factuality | Categorymath | Rank87 | Rating / score1,379.08 | Votes1,777 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry medicine and healthcare | Rank89 | Rating / score1,408.79 | Votes2,276 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry legal and government | Rank91 | Rating / score1,404.69 | Votes2,346 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryexpert | Rank96 | Rating / score1,408.12 | Votes2,723 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryrussian | Rank102 | Rating / score1,383.31 | Votes3,631 | ObservationsN/A | Result date |
| Arenatext factuality | Categorycreative writing | Rank105 | Rating / score1,354.16 | Votes4,780 | ObservationsN/A | Result date |
| Arenatext style control | Categorygerman | Rank107 | Rating / score1,389.65 | Votes644 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry life and physical and social science | Rank108 | Rating / score1,400.10 | Votes5,169 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry entertainment and sports and media | Rank109 | Rating / score1,352.09 | Votes6,479 | ObservationsN/A | Result date |
| Arenatext factuality | Categorymulti turn | Rank109 | Rating / score1,390.94 | Votes5,675 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry writing and literature and language | Rank113 | Rating / score1,369.62 | Votes7,750 | ObservationsN/A | Result date |
| Arenatext factuality | Categorylonger query | Rank113 | Rating / score1,397.23 | Votes13,952 | ObservationsN/A | Result date |
| Arenatext style control | Categorypolish | Rank113 | Rating / score1,371.38 | Votes928 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryinstruction following | Rank114 | Rating / score1,381.64 | Votes11,561 | ObservationsN/A | Result date |
| Arenatext style control | Categoryrussian | Rank114 | Rating / score1,394.54 | Votes5,227 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry business and management and financial operations | Rank115 | Rating / score1,384.48 | Votes6,259 | ObservationsN/A | Result date |
| Arenatext factuality | Categorycoding | Rank116 | Rating / score1,428.69 | Votes9,196 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryhard prompts english | Rank116 | Rating / score1,421.07 | Votes10,096 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryenglish | Rank117 | Rating / score1,403.81 | Votes18,283 | ObservationsN/A | Result date |
| Arenatext style control | Categoryexpert | Rank117 | Rating / score1,425.27 | Votes3,779 | ObservationsN/A | Result date |
| Arenatext style control | Categoryhard prompts english | Rank117 | Rating / score1,429.83 | Votes12,185 | ObservationsN/A | Result date |
MiniMax M2.5 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.30 input, $1.2 output per 1M
- Official provider
- MiniMax (minimax.io)
- Lowest third-party
- From $0.15 input, $1.15 output per 1M via Deep Infra
- Tracked offerings
- 45
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderAlibaba Token Plan | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context196.6K | Updated |
| ProviderMiniMax Token Plan (minimax.cn) | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context204.8K | Updated |
| ProviderAlibaba Coding Plan (China) | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context196.6K | Updated |
| ProviderMiniMax Token Plan (minimax.io) | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context204.8K | Updated |
| ProviderAlibaba Coding Plan | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context196.6K | Updated |
| ProviderTencent Coding Plan (China) | Provider model IDminimax-m2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context204.8K | Updated |
| ProviderSCNet Token Plan | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context204.8K | Updated |
| ProviderAlibaba Token Plan (China) | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context196.6K | Updated |
| ProviderDeep Infra | Provider model IDMiniMaxAI/MiniMax-M2.5 | Regionglobal | Input / 1M$0.15 | Output / 1M$1.15 | Context196.6K | Updated |
| ProviderDInference | Provider model IDminimax-m2.5 | Regionglobal | Input / 1M$0.22 | Output / 1M$0.88 | Context200K | Updated |
| ProviderNanoGPT | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.27 | Output / 1M$1.08 | Context204.8K | Updated |
| ProviderOpenRouter | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.27 | Output / 1M$0.95 | Context204.8K | Updated |
| ProviderKilo Gateway | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.27 | Output / 1M$0.95 | Context198K | Updated |
| ProviderVenice AI | Provider model IDminimax-m25 | Regionglobal | Input / 1M$0.27 | Output / 1M$0.95 | Context198K | Updated |
| ProviderD.Run (China) | Provider model IDpublic/minimax-m25 | Regionglobal | Input / 1M$0.29 | Output / 1M$1.16 | Context204.8K | Updated |
| ProviderHPC-AI | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context196K | Updated |
| ProviderMeganova | Provider model IDMiniMaxAI/MiniMax-M2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderTokenGo | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderOrcaRouter | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderHugging Face | Provider model IDMiniMaxAI/MiniMax-M2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderAlibaba (China) | Provider model IDMiniMax-M2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderAmazon Bedrock | Provider model IDminimax.minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context196.6K | Updated |
| ProviderLLM Gateway | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderCloudFerro Sherlock | Provider model IDMiniMaxAI/MiniMax-M2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context196K | Updated |
| ProviderMerge Gateway | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderDigitalOcean | Provider model IDminimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context65.5K | Updated |
| ProviderZenMux | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderTogether AI | Provider model IDMiniMaxAI/MiniMax-M2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderOfox | Provider model IDminimax/minimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
| ProviderOpenCode | Provider model IDminimax-m2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context204.8K | Updated |
MiniMax M2.5 Runtime Performance
Provider-specific output speed and catalog latency for MiniMax M2.5. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderMiniMax | Output Speed100.00 tok/s | Catalog Latency3.00 s | Max Input1M | Max Output1M | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
MiniMax M2.5 Specifications
Technical details for the model's default version.
- Version
- MiniMax M2.5
- Released
- Feb 12, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 230B
- Context window
- 1M
- Max output
- 1M
- Inputs
- image, text
- Outputs
- text
- Open weights
- Yes
- License
- MIT
MiniMax M2.5 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionMiniMax M2.5 | Released | LLMBoard53.25 | Parameters230B | Context1M | Max output1M | Open weightsYes | LicenseMIT |
Models similar to MiniMax M2.5
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
MI
MiniMax M2.1
MiniMax
MI
MiniMax M2.7
MiniMax
MI
MiniMax M2
MiniMax
MI
MiniMax M3
MiniMax
MI
MiniMax M1 80K
MiniMax
MI
MiniMax M1 40K
MiniMax
What is MiniMax M2.5?
Key information about MiniMax M2.5 and its available data.
1 foundation. It supports full-stack development for PC, App, and cross-platform applications, as well as agentic workflows such as automated customer support, data-analysis pipelines, and complex task execution.
Data as of 2026-10-22.
FAQ
Common questions about MiniMax M2.5.
When was MiniMax M2.5 released?
MiniMax M2.5's default version was released on Feb 12, 2026.
How much does MiniMax M2.5 cost?
MiniMax M2.5's official API price is $0.30 per million input tokens and $1.2 per million output tokens via MiniMax (minimax.io). The lowest tracked third-party offer starts at $0.15 input and $1.15 output via Deep Infra.
Who created MiniMax M2.5?
MiniMax M2.5 was created by MiniMax.
What is the context window for MiniMax M2.5?
The default version has a 1M token context window.
Is MiniMax M2.5 open weight?
Yes. The default version is marked as open weight under MIT.
How many API providers offer MiniMax M2.5?
45 provider offerings are linked to the default version.