DeepSeek V4 Pro 0813 vs GLM 5.3 Flash: Benchmark Comparison
DeepSeek V4 Pro 0813 has the higher score on 14 of 26 shared benchmarks; GLM 5.3 Flash leads on 11.
The largest observed score gap is 51.54 pts on Vibe Code Bench v1.1 , where DeepSeek V4 Pro 0813 leads.
Reported ±1 standard-error ranges overlap on 15 of 26 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 | GLM 5.3 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Finance Index | 52.94% ±1.41 | 55.83% ±1.51 | 2.89 pts | Reported ±1 SE ranges overlap |
| Vals Index | 52.97% ±1.04 | 50.86% ±1.25 | 2.11 pts | Reported ±1 SE ranges overlap |
| Vals Legal Index | 69.01% ±1.77 | 71.05% ±1.83 | 2.04 pts | Reported ±1 SE ranges overlap |
| HLAB-Vals-1.1 | 10.83% ±2.72 | 10.42% ±2.57 | 0.42 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 40.87% ±3.42 | 45.19% ±3.46 | 4.33 pts | Reported ±1 SE ranges overlap |
| LegalBench | 82.36% ±0.44 | 83.93% ±0.43 | 1.57 pts | Reported ±1 SE ranges do not overlap |
| EMB v1.1 | 65.13% ±2.35 | 66.25% ±2.69 | 1.12 pts | Reported ±1 SE ranges overlap |
| Finance Agent v2.1 | 57.12% ±2.08 | 60.59% ±2.03 | 3.47 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 26.76% ±3.09 | 31.55% ±3.35 | 4.79 pts | Reported ±1 SE ranges overlap |
| TaxEval v2 | 73.06% ±0.87 | 75.59% ±0.84 | 2.53 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 80.17% ±2.00 | 88.94% ±1.91 | 8.76 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 50.00% ±5.03 | 21.00% ±4.09 | 29.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 4.29% ±2.44 | 2.86% ±2.01 | 1.43 pts | Reported ±1 SE ranges overlap |
| GPQA Diamond | 92.42% ±2.02 | 86.36% ±2.13 | 6.06 pts | Reported ±1 SE ranges do not overlap |
| MMLU Pro | 86.97% ±0.34 | 86.06% ±0.34 | 0.91 pts | Reported ±1 SE ranges do not overlap |
| Code Migration v1.1 | 51.99% ±2.23 | 51.65% ±2.99 | 0.34 pts | Reported ±1 SE ranges overlap |
| IOI | 51.61% ±2.40 | 52.50% ±2.62 | 0.89 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 87.53% ±0.96 | 80.51% ±1.05 | 7.02 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.00% ±0.00 | 0.00% ±0.00 | 0.00 pts | Reported ±1 SE ranges overlap |
| SkillsBench | 53.83% ±4.49 | 40.16% ±4.43 | 13.67 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench | 96.40% ±0.83 | 92.00% ±1.21 | 4.40 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 14.14% ±2.02 | 37.37% ±2.67 | 23.23 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench 1-100 | 17.53% ±3.41 | 16.03% ±3.28 | 1.50 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 82.30% ±3.15 | 30.76% ±5.28 | 51.54 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 72.86% ±5.50 | 67.86% ±5.57 | 5.00 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 1.91% ±0.85 | 0.76% ±0.54 | 1.15 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | DeepSeek V4 Pro 0813 average | GLM 5.3 Flash average |
|---|---|---|
| Legal | 44.69% | 46.51% |
| Finance | 55.51% | 58.50% |
| Healthcare | 80.17% | 88.94% |
| Math | 50.00% | 21.00% |
| Science | 4.29% | 2.86% |
| Academic | 89.70% | 86.21% |
| Coding | 50.59% | 44.55% |
| Cyber | 37.38% | 34.31% |
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 | GLM 5.3 Flash cost | DeepSeek V4 Pro 0813 latency | GLM 5.3 Flash latency |
|---|---|---|---|---|
| Vals Finance Index | $0.88 | $0.31 | 24m00s | 33m49s |
| Vals Index | $2.50 | $1.13 | 58m03s | 1h04m |
| Vals Legal Index | $0.87 | $0.19 | 26m33s | 28m23s |
| HLAB-Vals-1.1 | $0.80 | $0.32 | 10m55s | 12m25s |
| Legal Research Bench | $1.09 | $0.12 | 42m07s | 51m07s |
| LegalBench | N/A | N/A | 20.05s | 12.52s |
| EMB v1.1 | $1.08 | $0.71 | 37m27s | 1h15m |
| Finance Agent v2.1 | $0.82 | $0.07 | 7m57s | 4m56s |
| Tax Agent Bench | $0.73 | $0.14 | 26m37s | 21m37s |
| TaxEval v2 | N/A | N/A | 3m46s | 53.06s |
| MedScribe | N/A | N/A | 2m36s | 86.98s |
| ProofBench v1.1 | $0.06 | $0.22 | 12m12s | 1h11m |
| Terminal-Bench-Science | $4.63 | $0.52 | 2h16m | 3h28m |
| GPQA Diamond | N/A | N/A | 4m39s | 2m11s |
| MMLU Pro | N/A | N/A | 59.62s | 44.76s |
| Code Migration v1.1 | $12.16 | $4.48 | 2h54m | 3h20m |
| IOI | $2.15 | $0.43 | 1h07m | 1h32m |
| LiveCodeBench | N/A | N/A | 4m33s | 5m02s |
| ProgramBench | $0.43 | $0.80 | 1h23m | 5h30m |
| SkillsBench | $0.37 | $0.04 | 11m15s | 12m09s |
| SWE-bench | $0.10 | $0.02 | 4m00s | 1h04m |
| Terminal-Bench 4.0 | $3.31 | $1.22 | 1h32m | 1h01m |
| Vibe Code Bench 1-100 | $5.89 | $0.46 | 37m42s | 39m03s |
| Vibe Code Bench v1.1 | $0.36 | $2.35 | 1h06m | 1h24m |
| CyberBench v1.1 | $0.59 | $0.14 | 22m33s | 35m48s |
| SRE Bench | $5.58 | $2.29 | 2h20m | 3h15m |
Results available only for DeepSeek V4 Pro 0813
- MedCode
Results available only for GLM 5.3 Flash
- MMMU Pro