MiniMax M2.5
Open Sourcevon minimax · Veroeffentlicht 2026-02-12
55.1
Durchschn. Score
$0.15/1M
Eingabepreis
$1.20/1M
Ausgabepreis
197K tokens (~98 books)
Kontextfenster
text
Typ
Tested on 21 benchmarks with 55.1% average. Top scores: Chatbot Arena Elo — Overall (1404.4%), Chatbot Arena Elo — Coding (1396.3%), OpenCompass — IFEval (91.1%).
Benchmark-Ergebnisse
| Benchmark | Kategorie | Score | Bar |
|---|---|---|---|
| Chatbot Arena Elo — Overall | arena | 1404.4 | |
| Chatbot Arena Elo — Coding | arena | 1396.3 | |
| OpenCompass — IFEval | language | 91.1 | |
| OpenCompass — AIME2025 | math | 86.2 | |
| OpenCompass — GPQA-Diamond | knowledge | 84.6 | |
| OpenCompass — MMLU-Pro | knowledge | 81.7 | |
| LiveBench — Mathematics | math | 77.4 | |
| OpenCompass — LiveCodeBenchV6 | coding | 73.6 | |
| LiveBench — Coding | coding | 70.7 | |
| ARC-AGI | reasoning | 63.7 | |
| LiveBench — Overall | knowledge | 60.1 | |
| LiveBench — Reasoning | reasoning | 59.3 | |
| LiveBench — If | language | 57.2 | |
| LiveBench — Language | language | 55.1 | |
| LiveBench — Agentic Coding | coding | 51.7 | |
| LiveBench — Data Analysis | reasoning | 49.6 | |
| Terminal Bench | coding | 42.2 | |
| OpenCompass — HLE | knowledge | 22.2 | |
| PostTrainBench | knowledge | 9.5 | |
| APEX-Agents | agentic | 6.2 | |
| ARC-AGI-2 | reasoning | 4.9 |
Aehnliche Modelle
DeepSeek
55.1
OpenAI
55.0
OpenAI
55.2
Alibaba Qwen
55.3