Claude Opus 5 vs Gemini 4 Argon: Benchmark Comparison
Claude Opus 5 has the higher score on 10 of 22 shared benchmarks; Gemini 4 Argon leads on 11.
The largest observed score gap is 33.67 pts on Time Horizon Index: KSP , where Gemini 4 Argon leads.
Reported ±1 standard-error ranges overlap on 8 of 20 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 | Claude Opus 5 | Gemini 4 Argon | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 33.02% | 30.55% | 2.47 pts | Uncertainty comparison unavailable |
| Vals Science Index | 47.93% ±2.20 | 55.36% ±2.29 | 7.43 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 55.29% ±3.46 | 54.81% ±3.46 | 0.48 pts | Reported ±1 SE ranges overlap |
| LegalBench | 86.97% ±0.42 | 88.30% ±0.37 | 1.33 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 45.55% ±3.59 | 44.56% ±3.59 | 0.99 pts | Reported ±1 SE ranges overlap |
| MedCode | 63.57% ±1.99 | 58.80% ±2.10 | 4.77 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 90.98% ±1.92 | 87.43% ±1.96 | 3.56 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 99.00% ±1.00 | 99.00% ±1.00 | 0.00 pts | Reported ±1 SE ranges overlap |
| BioMysteryBench | 79.26% ±2.06 | 76.30% ±0.37 | 2.96 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 37.39% ±3.25 | 45.49% ±3.35 | 8.11 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 27.14% ±5.35 | 44.29% ±5.98 | 17.14 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 49.43% ±3.29 | 53.65% ±3.38 | 4.22 pts | Reported ±1 SE ranges overlap |
| IOI | 84.33% ±9.96 | 100.00% ±0.00 | 15.67 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 3.00% ±1.21 | 2.50% ±1.11 | 0.50 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 53.53% ±1.34 | 57.58% ±2.31 | 4.04 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench 1-100 | 28.53% ±4.33 | 21.20% ±3.81 | 7.33 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 88.40% ±3.00 | 91.91% ±1.90 | 3.50 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 65.36% ±5.55 | 77.86% ±5.30 | 12.50 pts | Reported ±1 SE ranges do not overlap |
| SRE Bench | 12.21% ±2.03 | 44.27% ±3.08 | 32.06 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 76.93% ±1.10 | 69.76% ±0.00 | 7.17 pts | Reported ±1 SE ranges do not overlap |
| CUA-bench | 9.00% | 4.83% | 4.17 pts | Uncertainty comparison unavailable |
| Time Horizon Index: KSP | 18.83% ±0.00 | 52.50% ±8.86 | 33.67 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Opus 5 average | Gemini 4 Argon average |
|---|---|---|
| Legal | 71.13% | 71.55% |
| Finance | 45.55% | 44.56% |
| Healthcare | 77.28% | 73.11% |
| Math | 99.00% | 99.00% |
| Science | 47.93% | 55.36% |
| Education | 49.43% | 53.65% |
| Coding | 51.56% | 54.64% |
| Cyber | 38.79% | 61.07% |
| Social Mobility | 76.93% | 69.76% |
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 | Claude Opus 5 cost | Gemini 4 Argon cost | Claude Opus 5 latency | Gemini 4 Argon latency |
|---|---|---|---|---|
| Vals RSI Index | $418.69 | $179.80 | 90h00m | 90h00m |
| Vals Science Index | $13.04 | $18.40 | 1h02m | 57m00s |
| Legal Research Bench | $6.76 | $6.45 | 31m01s | 31m21s |
| LegalBench | N/A | N/A | 4.42s | 12.62s |
| Tax Agent Bench | $5.14 | $3.47 | 16m48s | 16m26s |
| MedCode | N/A | N/A | 24.50s | 79.53s |
| MedScribe | N/A | N/A | 76.56s | 87.44s |
| ProofBench v1.1 | $2.11 | $0.78 | 9m51s | 6m20s |
| BioMysteryBench | $3.29 | $5.71 | 36m36s | 39m57s |
| MysteryMechanism | $3.28 | $4.59 | 16m53s | 24m55s |
| Terminal-Bench-Science | $32.54 | $44.90 | 2h12m | 1h46m |
| SAGE | N/A | N/A | 56.17s | 63.07s |
| IOI | $16.48 | $5.31 | 56m42s | 24m19s |
| ProgramBench | $60.29 | $17.66 | 2h55m | 1h11m |
| Terminal-Bench 4.0 | $18.60 | $17.64 | 1h07m | 53m35s |
| Vibe Code Bench 1-100 | $41.49 | $18.66 | 1h22m | 1h43m |
| Vibe Code Bench v1.1 | $33.88 | $8.11 | 1h28m | 37m28s |
| CyberBench v1.1 | $2.87 | $5.11 | 19m25s | 19m10s |
| SRE Bench | $23.63 | $30.45 | 1h21m | 1h35m |
| Public Benefits Bench v1.1 | $3.95 | $2.79 | 30m09s | 26m07s |
| CUA-bench | $232.23 | $193.78 | 25.29s | 70.51s |
| Time Horizon Index: KSP | $1348.77 | $708.63 | N/A | N/A |
Results available only for Claude Opus 5
- MortgageTax
- TaxEval v2
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- LiveCodeBench
- SkillsBench
- SWE-bench
Results available only for Gemini 4 Argon
- Vals Finance Index
- Vals Index
- Vals Legal Index
- HLAB-Vals-1.1
- EMB v1.1
- Finance Agent v2.1
- Robo-bench
- Code Migration v1.1
- SWE-bench Multimodal