Qwen3 Next 80B A3B Thinking
Open Sourcevon Alibaba Qwen · Veroeffentlicht 2025-09-11
61.6
Durchschn. Score
$0.10/1M
Eingabepreis
$0.78/1M
Ausgabepreis
131K tokens (~66 books)
Kontextfenster
text
Typ
Tested on 20 benchmarks with 61.6% average. Top scores: Chatbot Arena Elo — Overall (1369.0%), OpenCompass — IFEval (89.5%), OpenCompass — AIME2025 (89.0%).
Benchmark-Ergebnisse
| Benchmark | Kategorie | Score | Bar |
|---|---|---|---|
| Chatbot Arena Elo — Overall | arena | 1369.0 | |
| OpenCompass — IFEval | language | 89.5 | |
| OpenCompass — AIME2025 | math | 89.0 | |
| OpenCompass — MMLU-Pro | knowledge | 82.0 | |
| HELM — IFEval | language | 81.0 | |
| HELM — WildBench | reasoning | 80.7 | |
| HELM — MMLU-Pro | knowledge | 78.6 | |
| OpenCompass — GPQA-Diamond | knowledge | 77.0 | |
| LiveBench — Mathematics | math | 74.3 | |
| OpenCompass — LiveCodeBenchV6 | coding | 66.3 | |
| HELM — GPQA | knowledge | 63.0 | |
| LiveBench — Coding | coding | 60.7 | |
| LiveBench — Reasoning | reasoning | 58.2 | |
| LiveBench — Language | language | 56.3 | |
| LiveBench — Data Analysis | reasoning | 53.6 | |
| LiveBench — Overall | knowledge | 50.4 | |
| HELM — Omni-MATH | math | 46.7 | |
| LiveBench — If | language | 41.5 | |
| OpenCompass — HLE | knowledge | 13.5 | |
| LiveBench — Agentic Coding | coding | 8.3 |