DeepSeek V4 Pro 0813 vs Gemini 3.6 Flash: Benchmark Comparison
DeepSeek V4 Pro 0813 has the higher score on 9 of 19 shared benchmarks; Gemini 3.6 Flash leads on 8.
The largest observed score gap is 27.50 pts on CyberBench v1.1 , where DeepSeek V4 Pro 0813 leads.
Reported ±1 standard-error ranges overlap on 5 of 19 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 | DeepSeek V4 Pro 0813 | Gemini 3.6 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | 7.50% ±2.15 | 3.33% ±1.17 | 4.17 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 40.87% ±3.42 | 25.00% ±3.01 | 15.87 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 82.36% ±0.44 | 86.70% ±0.41 | 4.34 pts | Reported ±1 SE ranges do not overlap |
| EMB | 52.80% ±3.06 | 65.41% ±2.65 | 12.61 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 50.39% ±0.25 | 56.30% ±0.18 | 5.90 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 58.66% ±3.14 | 49.90% ±3.30 | 8.77 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 73.06% ±0.87 | 74.86% ±0.85 | 1.80 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 42.47% ±2.16 | 53.15% ±2.16 | 10.68 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 80.17% ±2.00 | 79.66% ±1.86 | 0.51 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench Science | 4.29% ±2.44 | 4.29% ±2.44 | 0.00 pts | Reported ±1 SE ranges overlap |
| GPQA Diamond | 92.42% ±2.02 | 93.43% ±1.33 | 1.01 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 86.97% ±0.34 | 89.28% ±0.30 | 2.30 pts | Reported ±1 SE ranges do not overlap |
| Code Migration | 41.54% ±4.30 | 30.93% ±4.07 | 10.61 pts | Reported ±1 SE ranges do not overlap |
| IOI | 51.61% ±2.40 | 35.06% ±5.70 | 16.55 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 87.53% ±0.96 | 88.08% ±0.94 | 0.55 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 | 96.40% ±0.83 | 79.60% ±1.80 | 16.80 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 82.30% ±3.15 | 64.00% ±4.29 | 18.29 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 72.86% ±5.50 | 45.36% ±3.75 | 27.50 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | DeepSeek V4 Pro 0813 average | Gemini 3.6 Flash average |
|---|---|---|
| Legal | 43.58% | 38.35% |
| Finance | 58.73% | 61.61% |
| Healthcare | 61.32% | 66.41% |
| Science | 4.29% | 4.29% |
| Academic | 89.70% | 91.35% |
| Coding | 59.90% | 49.61% |
| Cyber | 72.86% | 45.36% |
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 | DeepSeek V4 Pro 0813 cost | Gemini 3.6 Flash cost | DeepSeek V4 Pro 0813 latency | Gemini 3.6 Flash latency |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | $0.17 | $1.90 | 21m53s | 16m24s |
| Legal Research Bench | $1.09 | $0.47 | 42m07s | 2m48s |
| LegalBench | N/A | N/A | 20.05s | 2.96s |
| EMB | $1.08 | $3.89 | 37m27s | 12m06s |
| Finance Agent (v2) | $0.88 | $1.40 | 16m15s | 3m45s |
| Tax Agent Bench | $0.73 | $0.36 | 26m37s | 105.32s |
| TaxEval v2 | N/A | N/A | 3m46s | 12.34s |
| MedCode | N/A | N/A | 3m01s | 16.68s |
| MedScribe | N/A | N/A | 2m36s | 33.55s |
| Terminal-Bench Science | $4.63 | $4.39 | 2h16m | 1h52m |
| GPQA Diamond | N/A | N/A | 4m39s | 18.06s |
| MMLU Pro | N/A | N/A | 59.62s | 7.30s |
| Code Migration | $18.58 | $8.93 | 3h32m | 1h30m |
| IOI | $2.15 | $8.67 | 1h07m | 48m23s |
| LiveCodeBench | N/A | N/A | 4m33s | 31.88s |
| ProgramBench | $0.43 | $5.92 | 1h23m | 47m33s |
| SWE-bench | $0.10 | $1.19 | 4m00s | 13m08s |
| Vibe Code Bench v1.1 | $0.36 | $3.04 | 1h06m | 25m23s |
| CyberBench v1.1 | $0.59 | $1.31 | 22m33s | 11m00s |
Results available only for DeepSeek V4 Pro 0813
- Vals Index
- ProofBench v1.1
- SkillsBench
- Terminal-Bench 4.0
- Vibe Code Bench 1-100
Results available only for Gemini 3.6 Flash
- MortgageTax
- BioMysteryBench
- MMMU Pro
- SAGE
- Public Benefits Bench v1.1