qwen2-72b-instruct: benchmark scores, price and ranking
qwen2-72b-instruct from Alibaba ranks #657 of 743 LLMs on the LLMs Tiger aggregate benchmark score (43.9), based on 5 benchmark results captured up to 2026-10-13.
- Organisation
- Alibaba
- Type
- Open-weight / local model
- Open weights
- Yes
- Released
- 2024-06-07 (source: epoch)
- Cheapest API price
- $0.900 in / $0.900 out per million tokens (via together_ai, observed 2026-10-13T20:47Z)
- Aggregate rank
- #657 of 743 (score 43.9)
Relative to other models, qwen2-72b-instruct is strongest on Aider Code Editing (#47 of 92) and weakest on TheAgentCompany (#13 of 13).
Across the 5 benchmarks where at least ten models are scored, qwen2-72b-instruct is in the top 10% on 0 and in the top half on 0.
Its cheapest listed input price, $0.900 per million tokens, is lower than that of 38% of the 427 models with a listed price.
qwen2-72b-instruct benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| Aider Code Editing | 55.6% | #47 of 92 | Aider Aider Code Editing leaderboard | 2026-10-13 |
| GPQA Diamond | 40.8% | #219 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| METR time horizon | 2 min | #28 of 31 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WeirdML | 11.3% | #102 of 109 | Epoch AI Benchmarking Hub | 2026-10-13 |
| TheAgentCompany | 1.1% | #13 of 13 | Epoch AI Benchmarking Hub | 2026-10-13 |
About qwen2-72b-instruct
qwen2-72b-instruct is an open-weight model that can run on your own hardware from Alibaba. It was released on 2024-06-07, according to Epoch AI. The cheapest API listing we track charges $0.900 per million input tokens and $0.900 per million output tokens, through together_ai. We hold 5 benchmark results for it across 4 areas, and none is in the top tenth of its board.
Where qwen2-72b-instruct stands
On our aggregate score, qwen2-72b-instruct is #657 of 743 ranked models, in the lower half. Its score is 43.9; the scale is centred on 50, so it sits 6.1 points below the typical model. That beats 12% of the ranked field. Among the 74 ranked Alibaba models it is #70. The next one up is qwen-turbo-2024-11-01 at 44.5. The next one down is qwen3-8b at 43.4.
qwen2-72b-instruct, benchmark by benchmark
qwen2-72b-instruct lands below the middle on Aider Code Editing: 55.6%, #47 of 92, with 50% of the field scoring higher. Claude 3.5-Sonnet-20241022 leads at 84.2%; the closest result above is GPT-4o mini at 55.6%. It covers Aider official benchmark. The median score on this board is also 55.6%. Source: Aider Aider Code Editing leaderboard, captured 2026-10-13.
qwen2-72b-instruct lands below the middle on GPQA Diamond: 40.8%, #219 of 275, with 79% of the field scoring higher. GPT-6 Astra (max) leads at 95.8%; the closest result above is Mellum2-12B-A2.5B-Instruct at 40.9%. It covers expert-level science questions. The board median is 71.7%, so qwen2-72b-instruct is 30.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
METR time horizon is a weak spot. qwen2-72b-instruct scores 2 min, #28 of 31, behind 87% of the models measured. The leader, Claude Mythos Preview (Early), reaches 1,045 min. It covers length of task (minutes) done at 50% success. The board median is 23 min, so qwen2-72b-instruct is 21 minutes below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
WeirdML is a weak spot. qwen2-72b-instruct scores 11.3%, #102 of 109, behind 93% of the models measured. The leader, GPT-6 Astra (max), reaches 93.3%. It covers unusual machine-learning tasks. The board median is 41.2%, so qwen2-72b-instruct is 29.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
TheAgentCompany is a weak spot. qwen2-72b-instruct scores 1.1%, #13 of 13, behind 92% of the models measured. The leader, gemini-2.5-flash-preview-09-2025, reaches 41.1%. It covers work tasks in a simulated software company. The board median is 8.6%, so qwen2-72b-instruct is 7.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
General knowledge and chat: weak, ahead of 8% of the field on average across 2 benchmarks. Its best board here is METR time horizon (#28 of 31) and its weakest is WeirdML (#102 of 109).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so qwen2-72b-instruct costs more than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on grok-build-0.1 ($1.00) and claude-haiku-4-5-20251001 ($1.00).
Release history
The Alibaba model released before it was Qwen2-72B on 2024-06-07, which scores 50.1. The next one was Qwen2-VL-72B-Instruct on 2024-08-29, at 47.6.
Reading these numbers
The 5 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (4), Aider Aider Code Editing leaderboard (1). 1 of them come from an official leaderboard and 4 from datasets or third-party runs. With 5 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #654 Yi 6B (44.0)
- #655 Mistral Small 3.2 (44.0)
- #656 Hermes-2-Theta-Llama-3-70B (44.0)
- #658 ollama/hermes3 (43.9)
- #659 Qwen2-7B (43.9)
- #660 gemini-1.5-pro-002 (43.9)
More from Alibaba
- Qwen 3.6 Max (Preview) (#25, 54.1)
- Qwen3.8 Max (xhigh) (#37, 53.2)
- Qwen3.7 Max (#71, 52.3)
- ollama/qwen2.5-coder:32b (#93, 51.9)
- Qwen3 235B A22B diff, no think, Alibaba API (#122, 51.5)
- Qwen2.5-Coder-14B-Instruct (#135, 51.3)
Compare it on the live leaderboard → · How the aggregate score works