Qwen3.8 Max (0902) (xhigh): benchmark scores, price and ranking
Qwen3.8 Max (0902) (xhigh) from Alibaba ranks #163 of 743 LLMs on the LLMs Tiger aggregate benchmark score (50.8), based on 8 benchmark results captured up to 2026-10-13.
- Organisation
- Alibaba
- Type
- Frontier model
- Open weights
- No
- Cheapest API price
- $1.65 in / $4.95 out per million tokens (via deepinfra, observed 2026-10-13T19:47Z)
- Aggregate rank
- #163 of 743 (score 50.8)
Relative to other models, Qwen3.8 Max (0902) (xhigh) is strongest on OTIS Mock AIME (#8 of 189) and weakest on Furniture Assembly (#28 of 28).
Across the 8 benchmarks where at least ten models are scored, Qwen3.8 Max (0902) (xhigh) is in the top 10% on 2 and in the top half on 6.
Its cheapest listed input price, $1.65 per million tokens, is lower than that of 25% of the 427 models with a listed price.
Qwen3.8 Max (0902) (xhigh) benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| OTIS Mock AIME | 100.0% | #8 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| GPQA Diamond | 92.3% | #21 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Chess Puzzles | 40.0% | #14 of 122 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SimpleQA Verified | 47.3% | #24 of 69 | Epoch AI Benchmarking Hub | 2026-10-13 |
| FrontierMath T1-3 | 65.6% | #28 of 68 | Epoch AI Benchmarking Hub | 2026-10-13 |
| FrontierMath T4 | 34.1% | #22 of 50 | Epoch AI Benchmarking Hub | 2026-10-13 |
| FrontierSWE | 17.8% | #16 of 19 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Furniture Assembly | 20.0% | #28 of 28 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Qwen3.8 Max (0902) (xhigh)
Qwen3.8 Max (0902) (xhigh) is a frontier model from Alibaba. The cheapest API listing we track charges $1.65 per million input tokens and $4.95 per million output tokens, through deepinfra. We hold 8 benchmark results for it across 3 areas, and 2 are in the top tenth of their boards.
Where Qwen3.8 Max (0902) (xhigh) stands
On our aggregate score, Qwen3.8 Max (0902) (xhigh) is #163 of 743 ranked models, in the top quarter. Its score is 50.8; the scale is centred on 50, so it sits 0.8 points above the typical model. That beats 78% of the ranked field. Among the 74 ranked Alibaba models it is #8. The next one up is Qwen3.5 397B-A17B at 51.1. The next one down is openrouter/qwen/qwen-2.5-coder-32b-instruct at 50.7.
Qwen3.8 Max (0902) (xhigh), benchmark by benchmark
On OTIS Mock AIME, Qwen3.8 Max (0902) (xhigh) scores 100.0% and ranks #8 of 189. Only 7 models score higher, led by Claude Fable 5.1 (max) at 100.0%, which is 0.0 points ahead. It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so Qwen3.8 Max (0902) (xhigh) is 42.2 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On GPQA Diamond, Qwen3.8 Max (0902) (xhigh) scores 92.3% and ranks #21 of 275. Only 20 models score higher, led by GPT-6 Astra (max) at 95.8%, which is 3.5 points ahead. It covers expert-level science questions. The board median is 71.7%, so Qwen3.8 Max (0902) (xhigh) is 20.6 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Qwen3.8 Max (0902) (xhigh) manages 40.0% on Chess Puzzles, good for #14 of 122. GPT-6 Astra (max) leads at 72.0%, 32.0 points further up, and 89% of models sit below. The results either side are GPT-5.6 Luna (max) (40.0%) and Kimi K3 (max) (39.0%). It covers chess tactics puzzles. The board median is 12.5%, so Qwen3.8 Max (0902) (xhigh) is 27.5 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
SimpleQA Verified puts Qwen3.8 Max (0902) (xhigh) at 47.3%, #24 of 69. That beats 65% of the models tested, but 28.7 points short of gemini-4-argon-high (76.0%). The results either side are GPT-5.1 (high) (48.0%) and Claude Opus 4.6 (max) (47.0%). It covers short factual questions. The board median is 41.0%, so Qwen3.8 Max (0902) (xhigh) is 6.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On FrontierMath T1-3, Qwen3.8 Max (0902) (xhigh) reaches 65.6%. It ranks #28 of 68, ahead of 59% of the field, and 28.1 points behind GPT-6 Astra (max) at 93.7%. The results either side are Claude Sonnet 5 (max) (65.6%) and DeepSeek V4 Pro 0813 (max) (64.6%). It covers research-level maths, tiers 1-3. The board median is 55.6%, so Qwen3.8 Max (0902) (xhigh) is 10.0 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Qwen3.8 Max (0902) (xhigh) manages 34.1% on FrontierMath T4, good for #22 of 50. GPT-6.1 Sol (max) leads at 100.0%, 65.9 points further up, and 56% of models sit below. The results either side are Qwen3.7 Max (34.1%) and Claude Opus 4.7 (max) (31.7%). It covers research-level maths, tier 4. The board median is 30.5%, so Qwen3.8 Max (0902) (xhigh) is 3.7 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
FrontierSWE is a weak spot. Qwen3.8 Max (0902) (xhigh) scores 17.8%, #16 of 19, behind 79% of the models measured. The leader, GPT-6 Astra (max), reaches 65.5%. It covers hard multi-hour software engineering. The board median is 29.5%, so Qwen3.8 Max (0902) (xhigh) is 11.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Furniture Assembly is a weak spot. Qwen3.8 Max (0902) (xhigh) scores 20.0%, #28 of 28, behind 96% of the models measured. The leader, Claude Opus 5.5 (max), reaches 83.3%. It covers planning a physical assembly. The board median is 41.2%, so Qwen3.8 Max (0902) (xhigh) is 21.2 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Reasoning, maths and science: strong, ahead of 76% of the field on average across 4 benchmarks. Its best board here is OTIS Mock AIME (#8 of 189) and its weakest is FrontierMath T4 (#22 of 50). General knowledge and chat: solid, ahead of 51% of the field on average across 3 benchmarks. Its best board here is Chess Puzzles (#14 of 122) and its weakest is Furniture Assembly (#28 of 28).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Qwen3.8 Max (0902) (xhigh) 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), Qwen3.7 Flash ($0.030, 50.5), GLM-5.2 (high) ($0.060, 51.6). Similar scores cost more on GPT-5.2 Codex ($1.75) and gpt-5.2-2025-12-11-high ($1.75).
Reading these numbers
The 8 scores on this page come from 1 source: Epoch AI Benchmarking Hub (8). No official leaderboard supplied any of them, so treat them as third-party measurements. The leaderboard also lists 2 other reasoning-effort settings of this model, which redirect here. This page shows the highest.
Qwen3.8 Max (0902) (xhigh) is also listed at 2 other reasoning-effort settings; this page shows the highest-effort row, which is the one the leaderboard keeps by default.
Models ranked nearby
- #160 MiniMax-M1-80k (50.9)
- #161 Optimus Alpha (50.9)
- #162 DeepSeek-V2.5 (50.9)
- #164 GPT-4 (0613) (50.8)
- #165 Claude Code + GLM-5.3 (50.8)
- #166 gpt-4-0314 (50.8)
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