qwen3-8b: benchmark scores, price and ranking
qwen3-8b from Alibaba ranks #678 of 743 LLMs on the LLMs Tiger aggregate benchmark score (43.4), based on 9 benchmark results captured up to 2026-10-13.
- Organisation
- Alibaba
- Type
- Open-weight / local model
- Open weights
- Yes
- Released
- 2025-04-28 (source: epoch)
- Cheapest API price
- $0.040 in / $0.140 out per million tokens (via llamagate, observed 2026-10-13T19:47Z)
- Aggregate rank
- #678 of 743 (score 43.4)
Relative to other models, qwen3-8b is strongest on Epoch ECI (#67 of 204) and weakest on CritPt (#113 of 115).
Across the 9 benchmarks where at least ten models are scored, qwen3-8b is in the top 10% on 0 and in the top half on 2.
Its cheapest listed input price, $0.040 per million tokens, is lower than that of 95% of the 427 models with a listed price.
qwen3-8b benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| Epoch ECI | 136 | #67 of 204 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Fiction.liveBench | 62.1% | #23 of 47 | Epoch AI Benchmarking Hub | 2026-10-13 |
| OTIS Mock AIME | 56.1% | #97 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| GPQA Diamond | 56.8% | #172 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Chess Puzzles | 5.0% | #81 of 122 | Epoch AI Benchmarking Hub | 2026-10-13 |
| DTBench | 59.7% | #105 of 146 | Epoch AI Benchmarking Hub | 2026-10-13 |
| LMCA | 8.8% | #99 of 108 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SciCode | 22.6% | #101 of 106 | Epoch AI Benchmarking Hub | 2026-10-13 |
| CritPt | 0.0% | #113 of 115 | Epoch AI Benchmarking Hub | 2026-10-13 |
About qwen3-8b
qwen3-8b is an open-weight model that can run on your own hardware from Alibaba. It was released on 2025-04-28, according to Epoch AI. The cheapest API listing we track charges $0.040 per million input tokens and $0.140 per million output tokens, through llamagate. We hold 9 benchmark results for it across 4 areas, and none is in the top tenth of its board.
Where qwen3-8b stands
On our aggregate score, qwen3-8b is #678 of 743 ranked models, in the lower half. Its score is 43.4; the scale is centred on 50, so it sits 6.6 points below the typical model. That beats 9% of the ranked field. Among the 74 ranked Alibaba models it is #71. The next one up is qwen2-72b-instruct at 43.9. The next one down is Qwen2.5-Coder-0.5B-Instruct at 42.6.
qwen3-8b, benchmark by benchmark
Epoch ECI puts qwen3-8b at 136, #67 of 204. That beats 67% of the models tested, but 20 points short of Qwen 3.8 Max (156). The results either side are Qwen3-30B-A3B (136) and DeepSeek-V3 (Mar 2025) (136). It covers Epoch Capabilities Index: Epoch's own aggregate of ~40 benchmarks. The board median is 127, so qwen3-8b is 9 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On Fiction.liveBench, qwen3-8b reaches 62.1%. It ranks #23 of 47, ahead of 51% of the field, and 32.3 points behind grok-4-0709 at 94.4%. The results either side are qwen3-14b (62.5%) and Claude Opus 4 (61.1%). It covers long-context story comprehension (16k). The board median is 61.1%, so qwen3-8b is 1.0 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
qwen3-8b lands below the middle on OTIS Mock AIME: 56.1%, #97 of 189, with 51% of the field scoring higher. Claude Fable 5.1 (max) leads at 100.0%; the closest result above is Gemini 2.0 Flash Thinking Exp at 57.8%. It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so qwen3-8b is 1.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
qwen3-8b lands below the middle on GPQA Diamond: 56.8%, #172 of 275, with 62% of the field scoring higher. GPT-6 Astra (max) leads at 95.8%; the closest result above is Gemini 2.0 Flash Thinking Exp at 57.1%. It covers expert-level science questions. The board median is 71.7%, so qwen3-8b is 15.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
qwen3-8b lands below the middle on Chess Puzzles: 5.0%, #81 of 122, with 66% of the field scoring higher. GPT-6 Astra (max) leads at 72.0%; the closest result above is Qwen3-32B at 5.0%. It covers chess tactics puzzles. The board median is 12.5%, so qwen3-8b is 7.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
qwen3-8b lands below the middle on DTBench: 59.7%, #105 of 146, with 71% of the field scoring higher. Claude Opus 5.5 (max) leads at 98.9%; the closest result above is Mistral Small 3.2 at 59.9%. It covers dynamic tool-use benchmark. The board median is 68.8%, so qwen3-8b is 9.1 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
LMCA is a weak spot. qwen3-8b scores 8.8%, #99 of 108, behind 91% of the models measured. The leader, Claude Opus 5.5 (max), reaches 68.2%. It covers language-model capability assessment. The board median is 29.2%, so qwen3-8b is 20.4 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
SciCode is a weak spot. qwen3-8b scores 22.6%, #101 of 106, behind 94% of the models measured. The leader, Claude Opus 5.5 (max), reaches 66.9%. It covers scientific research code. The board median is 44.3%, so qwen3-8b is 21.8 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
CritPt is a weak spot. qwen3-8b scores 0.0%, #113 of 115, behind 97% of the models measured. The leader, GPT-5.6 Sol (max), reaches 32.3%. It covers unpublished research-level physics problems. The board median is 2.9%, so qwen3-8b is 2.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Reasoning, maths and science: weak, ahead of 29% of the field on average across 3 benchmarks. Its best board here is OTIS Mock AIME (#97 of 189) and its weakest is CritPt (#113 of 115). General knowledge and chat: uneven, ahead of 40% of the field on average across 4 benchmarks. Its best board here is Epoch ECI (#67 of 204) and its weakest is LMCA (#99 of 108).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so qwen3-8b costs less 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 Mistral-7B-Instruct-v0.2 ($0.050) and qwen-turbo-2024-11-01 ($0.050).
Release history
The Alibaba model released before it was Qwen3-30B-A3B on 2025-04-28, which scores 44.9. The next one was qwen3-1.7b on 2025-04-29, at 46.5.
Reading these numbers
The 9 scores on this page come from 1 source: Epoch AI Benchmarking Hub (9). No official leaderboard supplied any of them, so treat them as third-party measurements. The leaderboard also lists 1 other reasoning-effort setting of this model, which redirect here. This page shows the highest.
qwen3-8b is also listed at 1 other reasoning-effort setting; this page shows the highest-effort row, which is the one the leaderboard keeps by default.
Models ranked nearby
- #675 Qwen2-1.5B-Instruct (43.4)
- #676 Qwen2-1.5B (43.4)
- #677 grok-2-1212 (43.4)
- #679 Llama-3.2-3B (43.4)
- #680 mistral-large-2407 (43.3)
- #681 granite-3.0-2b-base (43.3)
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