Qwen 3.8 Max vs Gemini 3.6 Flash: Benchmark Comparison
Qwen 3.8 Max has the higher score on 12 of 23 shared benchmarks; Gemini 3.6 Flash leads on 10.
The largest observed score gap is 33.83 pts on IOI , where Qwen 3.8 Max leads.
Reported ±1 standard-error ranges overlap on 9 of 23 shared benchmarks with comparable uncertainty data. This is not a pairwise statistical significance test.
Shared benchmark results
Scores and reported standard errors come from the latest visible Vals benchmark version available for both models.
| Benchmark | Qwen 3.8 Max | Gemini 3.6 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | 10.42% ±2.29 | 3.33% ±1.17 | 7.08 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 47.60% ±3.47 | 25.00% ±3.01 | 22.60 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 83.61% ±0.42 | 86.70% ±0.41 | 3.10 pts | Reported ±1 SE ranges do not overlap |
| EMB | 60.07% ±3.35 | 65.41% ±2.65 | 5.34 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 50.59% ±0.58 | 56.30% ±0.18 | 5.70 pts | Reported ±1 SE ranges do not overlap |
| MortgageTax | 63.99% ±0.94 | 67.61% ±0.92 | 3.62 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 65.96% ±3.18 | 49.90% ±3.30 | 16.06 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 75.55% ±0.84 | 74.86% ±0.85 | 0.70 pts | Reported ±1 SE ranges overlap |
| MedCode | 40.67% ±2.03 | 53.15% ±2.16 | 12.48 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 84.95% ±2.00 | 79.66% ±1.86 | 5.29 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench Science | 12.86% ±4.03 | 4.29% ±2.44 | 8.57 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 93.69% ±1.22 | 93.43% ±1.33 | 0.25 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 88.60% ±0.31 | 89.28% ±0.30 | 0.67 pts | Reported ±1 SE ranges do not overlap |
| MMMU Pro | 88.03% ±0.78 | 88.38% ±0.77 | 0.35 pts | Reported ±1 SE ranges overlap |
| SAGE | 51.25% ±3.42 | 48.69% ±3.38 | 2.57 pts | Reported ±1 SE ranges overlap |
| Code Migration | 23.96% ±4.31 | 30.93% ±4.07 | 6.97 pts | Reported ±1 SE ranges overlap |
| IOI | 68.89% ±5.20 | 35.06% ±5.70 | 33.83 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 87.85% ±0.95 | 88.08% ±0.94 | 0.23 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 0.00% ±0.00 | 0.00% ±0.00 | 0.00 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 85.60% ±1.57 | 79.60% ±1.80 | 6.00 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 64.70% ±5.36 | 64.00% ±4.29 | 0.69 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 28.57% ±3.31 | 45.36% ±3.75 | 16.79 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 67.12% ±1.22 | 56.56% ±1.29 | 10.55 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Qwen 3.8 Max average | Gemini 3.6 Flash average |
|---|---|---|
| Legal | 47.21% | 38.35% |
| Finance | 63.23% | 62.81% |
| Healthcare | 62.81% | 66.41% |
| Science | 12.86% | 4.29% |
| Academic | 90.11% | 90.36% |
| Education | 51.25% | 48.69% |
| Coding | 55.17% | 49.61% |
| Cyber | 28.57% | 45.36% |
| Social Mobility | 67.12% | 56.56% |
Cost and latency
Cost per test appears only where the benchmark reports it for both models. Latency is the measured completion time for that benchmark.
| Benchmark | Qwen 3.8 Max cost | Gemini 3.6 Flash cost | Qwen 3.8 Max latency | Gemini 3.6 Flash latency |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | $2.37 | $1.90 | 49m12s | 16m24s |
| Legal Research Bench | $2.49 | $0.47 | 1h01m | 2m48s |
| LegalBench | N/A | N/A | 36.37s | 2.96s |
| EMB | $2.66 | $3.89 | 57m47s | 12m06s |
| Finance Agent (v2) | $1.24 | $1.40 | 19m35s | 3m45s |
| MortgageTax | N/A | N/A | 47.50s | 9.24s |
| Tax Agent Bench | $1.55 | $0.36 | 40m36s | 105.32s |
| TaxEval v2 | N/A | N/A | 4m26s | 12.34s |
| MedCode | N/A | N/A | 6m18s | 16.68s |
| MedScribe | N/A | N/A | 4m11s | 33.55s |
| Terminal-Bench Science | $16.22 | $4.39 | 3h20m | 1h52m |
| GPQA Diamond | N/A | N/A | 4m27s | 18.06s |
| MMLU Pro | N/A | N/A | 63.58s | 7.30s |
| MMMU Pro | N/A | N/A | 43.85s | 11.31s |
| SAGE | N/A | N/A | 2m13s | 21.96s |
| Code Migration | $10.54 | $8.93 | 4h28m | 1h30m |
| IOI | $9.47 | $8.67 | 1h47m | 48m23s |
| LiveCodeBench | N/A | N/A | 5m38s | 31.88s |
| ProgramBench | $12.36 | $5.92 | 5h54m | 47m33s |
| SWE-bench | $1.13 | $1.19 | 41m42s | 13m08s |
| Vibe Code Bench v1.1 | $8.24 | $3.04 | 2h49m | 25m23s |
| CyberBench v1.1 | $0.78 | $1.31 | 9m40s | 11m00s |
| Public Benefits Bench v1.1 | $0.99 | $0.39 | 40m44s | 2m02s |
Results available only for Qwen 3.8 Max
- Vals Index
- Vals RSI Index
- ProofBench v1.1
- MysteryMechanism
- SkillsBench
- Terminal-Bench 4.0
- Vibe Code Bench 1-100
Results available only for Gemini 3.6 Flash
- BioMysteryBench