# Qwen2.5 72B Instruct

As of October 22, 2026, Qwen2.5 72B Instruct ranks #113 of 123 models overall with 26.5 ±10.5 on the LLMBenchmark Index. Qwen2.5 72B Instruct ranks #102 of 122 in reasoning, its strongest category. Qwen2.5 72B Instruct costs $0.36 per million input tokens and $0.40 per million output tokens. Qwen2.5 72B Instruct was released on September 19, 2024.

Page: https://llmbenchmark.ai/models/qwen-2-5-72b-instruct
Data from October 22, 2026. Every figure below names the board it was read from and the day.

- Maker: Qwen
- Released: September 19, 2024
- Price per million input tokens: $0.36
- Price per million output tokens: $0.40
- Context window: 33K tokens
- Overall rank: 113 of 123, by the LLMBenchmark Index
- LLMBenchmark Index: 26.5 ±10.5 (90% margin, from 4 of 7 inputs)
- Artificial Analysis Intelligence Index: 7.7

Benchmark results (53 of 195):

- AA Intelligence Index: 7.7 (source: Artificial Analysis (via OpenRouter), read October 22, 2026)
- AA Math Index: 14.0 (source: Artificial Analysis, read October 22, 2026)
- GPQA Diamond (Artificial Analysis run): 49.1% (source: Artificial Analysis, read October 22, 2026)
- Humanity's Last Exam (Artificial Analysis run): 3.6% (source: Artificial Analysis, read October 22, 2026)
- MMLU Pro (Artificial Analysis run): 72.0% (source: Artificial Analysis, read October 22, 2026)
- LiveCodeBench (Artificial Analysis run): 27.6% (source: Artificial Analysis, read October 22, 2026)
- Terminal-Bench Hard (Artificial Analysis run): 4.5% (source: Artificial Analysis, read October 22, 2026)
- IFBench: 36.9% (source: Artificial Analysis, read October 22, 2026)
- τ²-Bench Telecom: 34.5% (source: Artificial Analysis, read October 22, 2026)
- AIME 2025 (Artificial Analysis run): 14.0% (source: Artificial Analysis, read October 22, 2026)
- WeirdML: 16.0% (source: Epoch AI benchmarking hub, read October 22, 2026)
- METR time horizons: 35.8% (source: Epoch AI benchmarking hub, read October 22, 2026)
- BALROG: 16.2% (source: Epoch AI benchmarking hub, read October 22, 2026)
- GPQA Diamond (Epoch AI run): 49.1% (source: Epoch AI benchmarking hub, read October 22, 2026)
- OTIS Mock AIME: 8.1% (source: Epoch AI benchmarking hub, read October 22, 2026)
- MATH level 5: 63.2% (source: Epoch AI benchmarking hub, read October 22, 2026)
- MMLU: 85.3% (source: Epoch AI benchmarking hub, read October 22, 2026)
- DTBench: 62.9% (source: Epoch AI benchmarking hub, read October 22, 2026)
- LMCA: 13.4% (source: Epoch AI benchmarking hub, read October 22, 2026)
- TheAgentCompany: 5.7% (source: Epoch AI benchmarking hub, read October 22, 2026)
- BIG-Bench Hard: 79.8% (source: Epoch AI benchmarking hub, read October 22, 2026)
- AI2 Reasoning Challenge: 94.5% (source: Epoch AI benchmarking hub, read October 22, 2026)
- HellaSwag: 84.8% (source: Epoch AI benchmarking hub, read October 22, 2026)
- WinoGrande: 82.3% (source: Epoch AI benchmarking hub, read October 22, 2026)
- PIQA: 82.6% (source: Epoch AI benchmarking hub, read October 22, 2026)
- Epoch Capabilities Index: 129.0 (source: Epoch AI benchmarking hub, read October 22, 2026)
- LMArena Text: 1269 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: coding: 1292 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: hard prompts: 1271 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: math: 1283 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: instruction following: 1254 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: creative writing: 1221 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: English: 1283 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: non-English: 1252 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Chinese: 1273 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: German: 1234 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Japanese: 1180 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Korean: 1188 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: French: 1280 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Spanish: 1255 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Russian: 1264 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: multi-turn: 1272 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: longer queries: 1281 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: expert: 1245 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: hard prompts in English: 1279 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: software and IT: 1292 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: mathematical work: 1290 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: science: 1275 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: medicine and healthcare: 1258 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: legal and government: 1288 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: business and finance: 1253 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: writing and language: 1247 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: entertainment and media: 1211 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)