Inference
EN
MiniMax

MiniMax M2.5

MiniMax M2.5 is a standard model from MiniMax. 10 benchmarks count toward its index, in 5 categories.

  • Open weights
  • Standard
  • Input: text
  • Released 2026-02
RankedEnough results for an index.
IndexAll categories, out of 100.50.2±7.1Median 51.0
CoverageIndex weight that has results.80%
SpeedOutput tokens per second.44/sMedian 67/s
First tokenSeconds to the first output token.0.7 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.27Median $0.60
Output / 1MUS dollars per 1M output tokens.$1.08Median $2.55
ContextMaximum tokens in one request.200KMedian 262K
EloLMArena rating and rank.1359±4 #203Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
minimax/minimax-m2.5
Max outputMaximum output tokens per request.
128K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.027
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
6 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
OpenRouterRouter$0.27$1.08200K
Deep InfraDeprecatedHost$0.15$1.15197K
Amazon BedrockCloud$0.30$1.20197K
Novita AIHost$0.30$1.20205K
SiliconFlowDeprecatedHost$0.30$1.20197K
Together AIDeprecatedHost$0.30$1.20205K

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Run locally

228.69B parameters
16-bit
481.2 GB
8-bit
256.1 GB
4-bit
146.6 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.1 counted51.6—
CodingCode writing and repair.4 counted50.7133 / 209
ReasoningLogic problems and puzzles.2 counted48.8124 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.2 counted48.9142 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.1 counted54.3—

Results

10 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
SWE-bench Verified · SWE-bench/experimentsCoding75.8%3 / 65
AIMEMath88.8% ±0.928 / 95
Terminal-Bench 2.0 · Vals AIAgentic41.6% ±5.329 / 66
GPQA Diamond · Vals AIKnowledge82.1% ±2.360 / 135
LiveCodeBenchCoding79.2% ±1.172 / 142
ARC-AGI-1 (semi-private)Reasoning63.7%64 / 126
SWE-benchCoding74.2% ±2.049 / 87
ARC-AGI-2 (semi-private)Reasoning4.9%80 / 131
IOI v1Coding6.7% ±2.640 / 61
MMLU ProKnowledge80.1% ±0.489 / 135
Vibe Code Bench v1.1Coding14.9% ±3.098 / 123
MedQAHealthcare92.5% ±0.230 / 95
SWE-bench Multilingual · swebench.comMultilingual68.3%6 / 12
ARC-AGI-2 (public eval)Reasoning5.4%65 / 122
CorpFin v2Finance59.6% ±1.076 / 131
ARC-AGI-1 (public eval)Reasoning59.1%69 / 116
LegalBenchLegal80.0% ±0.591 / 149
TaxEval v2Finance68.2% ±0.9100 / 143
CaseLaw v2Legal53.5% ±0.539 / 53
Finance Agent v1.1Finance38.6% ±2.641 / 51

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence statedSWE-bench teamNo licence statedARC Prize FoundationTerms restrict reusemodels.devMIT

History

5 releases
Index51.2 → 50.2
23 Oct 202612 Nov 2026
Output / 1M$0.95 → $1.08
23 Oct 202612 Nov 2026

Questions

How much does MiniMax M2.5 cost?
$0.27 per 1M input tokens, $1.08 per 1M output tokens.
How good is MiniMax M2.5?
Its index is 50.2 out of 100. That is place 149 of 283 ranked models.
What is MiniMax M2.5 best at?
Knowledge: 48.9 out of 100, place 142 of 268.
How fast is MiniMax M2.5?
44 output tokens per second, 0.7 s to the first token.
What is the context window of MiniMax M2.5?
200K tokens.
Is MiniMax M2.5 open source?
Yes. Its weights are published.
When was MiniMax M2.5 released?
February 2026.

Badge

MiniMax M2.5 on Inference360: place 149 of 283, index 50.2
HTML
<a href="https://inference360.ai/models/minimax-m2-5"><img src="https://inference360.ai/badges/minimax-m2-5.svg" alt="MiniMax M2.5 on Inference360: place 149 of 283, index 50.2" height="20"></a>
Markdown
[![MiniMax M2.5 on Inference360: place 149 of 283, index 50.2](https://inference360.ai/badges/minimax-m2-5.svg)](https://inference360.ai/models/minimax-m2-5)

Compare with

GLM-4.750.0Grok 450.0MiniMax M2.149.1MiniMax M2.751.2

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from MiniMax

All models
MiniMax M2-herUnrankedMiniMax M2.751.2MiniMax M2.149.1MiniMax M355.7MiniMax M2Unranked