Qwen 3.8 Max vs Gemini 3.5 Flash: Benchmark Comparison
Qwen 3.8 Max has the higher score on 16 of 25 shared benchmarks; Gemini 3.5 Flash leads on 8.
The largest observed score gap is 28.28 pts on Terminal-Bench 4.0 , where Qwen 3.8 Max leads.
Reported ±1 standard-error ranges overlap on 9 of 25 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 3.5 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 48.27% ±1.34 | 44.79% ±1.10 | 3.48 pts | Reported ±1 SE ranges do not overlap |
| Harvey's Legal Agent Benchmark | 10.42% ±2.29 | 2.50% ±0.83 | 7.92 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 47.60% ±3.47 | 30.77% ±3.21 | 16.83 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 83.61% ±0.42 | 83.60% ±0.85 | 0.01 pts | Reported ±1 SE ranges overlap |
| EMB | 60.07% ±3.35 | 63.55% ±2.75 | 3.48 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 50.59% ±0.58 | 57.86% ±0.23 | 7.27 pts | Reported ±1 SE ranges do not overlap |
| MortgageTax | 63.99% ±0.94 | 68.12% ±0.91 | 4.13 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 65.96% ±3.18 | 53.53% ±3.30 | 12.43 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 75.55% ±0.84 | 74.37% ±0.85 | 1.19 pts | Reported ±1 SE ranges overlap |
| MedCode | 40.67% ±2.03 | 55.83% ±2.11 | 15.16 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 84.95% ±2.00 | 76.57% ±1.92 | 8.37 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 58.00% ±4.96 | 31.00% ±4.65 | 27.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench Science | 12.86% ±4.03 | 5.71% ±2.79 | 7.14 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 93.69% ±1.22 | 92.68% ±1.46 | 1.01 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 88.60% ±0.31 | 89.52% ±0.31 | 0.91 pts | Reported ±1 SE ranges do not overlap |
| MMMU Pro | 88.03% ±0.78 | 88.27% ±0.77 | 0.23 pts | Reported ±1 SE ranges overlap |
| SAGE | 51.25% ±3.42 | 49.88% ±3.44 | 1.37 pts | Reported ±1 SE ranges overlap |
| Code Migration | 23.96% ±4.31 | 26.75% ±4.10 | 2.79 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 87.85% ±0.95 | 87.60% ±0.95 | 0.25 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 0.00% ±0.00 | 0.00% ±0.00 | 0.00 pts | Reported ±1 SE ranges overlap |
| SkillsBench | 42.01% ±4.43 | 52.74% ±4.47 | 10.73 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench | 85.60% ±1.57 | 78.80% ±1.83 | 6.80 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 34.34% ±3.94 | 6.06% ±1.51 | 28.28 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 64.70% ±5.36 | 48.68% ±4.73 | 16.01 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 67.12% ±1.22 | 59.47% ±1.28 | 7.65 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Qwen 3.8 Max average | Gemini 3.5 Flash average |
|---|---|---|
| Legal | 47.21% | 38.96% |
| Finance | 63.23% | 63.48% |
| Healthcare | 62.81% | 66.20% |
| Math | 58.00% | 31.00% |
| Science | 12.86% | 5.71% |
| Academic | 90.11% | 90.15% |
| Education | 51.25% | 49.88% |
| Coding | 48.35% | 42.95% |
| Social Mobility | 67.12% | 59.47% |
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 3.5 Flash cost | Qwen 3.8 Max latency | Gemini 3.5 Flash latency |
|---|---|---|---|---|
| Vals Index | $4.96 | $3.30 | 1h41m | 18m17s |
| Harvey's Legal Agent Benchmark | $2.37 | $2.12 | 49m12s | 7m58s |
| Legal Research Bench | $2.49 | $1.31 | 1h01m | 4m38s |
| LegalBench | N/A | N/A | 36.37s | 34.58s |
| EMB | $2.66 | $5.17 | 57m47s | 18m44s |
| Finance Agent (v2) | $1.24 | $2.51 | 19m35s | 5m22s |
| MortgageTax | N/A | N/A | 47.50s | 11.43s |
| Tax Agent Bench | $1.55 | $0.86 | 40m36s | 3m53s |
| TaxEval v2 | N/A | N/A | 4m26s | 19.69s |
| MedCode | N/A | N/A | 6m18s | 25.29s |
| MedScribe | N/A | N/A | 4m11s | 57.08s |
| ProofBench v1.1 | $1.50 | $0.57 | 31m08s | 6m34s |
| Terminal-Bench Science | $16.22 | $3.45 | 3h20m | 59m32s |
| GPQA Diamond | N/A | N/A | 4m27s | 27.66s |
| MMLU Pro | N/A | N/A | 63.58s | 26.26s |
| MMMU Pro | N/A | N/A | 43.85s | 12.23s |
| SAGE | N/A | N/A | 2m13s | 32.59s |
| Code Migration | $10.54 | $5.73 | 4h28m | 19m28s |
| LiveCodeBench | N/A | N/A | 5m38s | 64.37s |
| ProgramBench | N/A | N/A | 5h54m | 18m25s |
| SkillsBench | $0.52 | $1.25 | 15m48s | 7m15s |
| SWE-bench | $1.13 | $0.95 | 41m42s | 4m14s |
| Terminal-Bench 4.0 | $10.63 | $5.98 | 2h28m | 1h11m |
| Vibe Code Bench v1.1 | $8.24 | $2.54 | 2h49m | 14m53s |
| Public Benefits Bench v1.1 | $0.99 | $1.11 | 40m44s | 5m39s |
Results available only for Qwen 3.8 Max
- Vals RSI Index
- MysteryMechanism
- IOI
- Vibe Code Bench 1-100
- CyberBench v1.1
Results available only for Gemini 3.5 Flash
None.