Gemini 4 Argon vs GPT-6.1 Sol: Benchmark Comparison
Gemini 4 Argon has the higher score on 18 of 27 shared benchmarks; GPT-6.1 Sol leads on 8.
The largest observed score gap is 38.57 pts on CyberBench v1.1 , where Gemini 4 Argon leads.
Reported ±1 standard-error ranges overlap on 11 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 | Gemini 4 Argon | GPT-6.1 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Finance Index | 64.89% ±1.49 | 55.81% ±1.50 | 9.08 pts | Reported ±1 SE ranges do not overlap |
| Vals Index | 69.24% ±1.09 | 63.94% ±1.07 | 5.29 pts | Reported ±1 SE ranges do not overlap |
| Vals Legal Index | 76.72% ±1.84 | 68.72% ±1.77 | 8.00 pts | Reported ±1 SE ranges do not overlap |
| Vals RSI Index | 30.55% | 26.62% | 3.93 pts | Uncertainty comparison unavailable |
| Vals Science Index | 55.36% ±2.29 | 59.63% ±2.31 | 4.27 pts | Reported ±1 SE ranges overlap |
| HLAB-Vals-1.1 | 27.08% ±3.74 | 6.25% ±1.83 | 20.83 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 54.81% ±3.46 | 38.46% ±3.38 | 16.35 pts | Reported ±1 SE ranges do not overlap |
| EMB v1.1 | 74.83% ±2.49 | 71.09% ±2.75 | 3.74 pts | Reported ±1 SE ranges overlap |
| Finance Agent v2.1 | 67.66% ±2.00 | 61.82% ±2.11 | 5.84 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 44.56% ±3.59 | 21.75% ±2.98 | 22.81 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 58.80% ±2.10 | 48.84% ±2.12 | 9.96 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 87.43% ±1.96 | 86.45% ±1.90 | 0.97 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 | 76.30% ±0.37 | 79.63% ±0.74 | 3.33 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 45.49% ±3.35 | 46.40% ±3.35 | 0.90 pts | Reported ±1 SE ranges overlap |
| Robo-bench | 0.83% | 28.33% | 27.50 pts | Uncertainty comparison unavailable |
| Terminal-Bench-Science | 44.29% ±5.98 | 52.86% ±6.01 | 8.57 pts | Reported ±1 SE ranges overlap |
| SAGE | 53.65% ±3.38 | 46.54% ±3.46 | 7.11 pts | Reported ±1 SE ranges do not overlap |
| Code Migration v1.1 | 70.80% ±3.72 | 77.32% ±2.98 | 6.53 pts | Reported ±1 SE ranges overlap |
| IOI | 100.00% ±0.00 | 96.89% ±3.11 | 3.11 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 2.50% ±1.11 | 3.00% ±1.21 | 0.50 pts | Reported ±1 SE ranges overlap |
| SWE-bench Multimodal | 56.04% ±2.27 | 43.54% ±2.27 | 12.50 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 57.58% ±2.31 | 55.05% ±1.82 | 2.52 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 91.91% ±1.90 | 88.93% ±1.89 | 2.98 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 77.86% ±5.30 | 39.29% ±2.74 | 38.57 pts | Reported ±1 SE ranges do not overlap |
| SRE Bench | 44.27% ±3.08 | 50.76% ±3.10 | 6.49 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 69.76% ±0.00 | 59.34% ±1.28 | 10.42 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Gemini 4 Argon average | GPT-6.1 Sol average |
|---|---|---|
| Legal | 40.95% | 22.36% |
| Finance | 62.35% | 51.55% |
| Healthcare | 73.11% | 67.65% |
| Math | 99.00% | 99.00% |
| Science | 41.73% | 51.80% |
| Education | 53.65% | 46.54% |
| Coding | 63.14% | 60.79% |
| Cyber | 61.07% | 45.02% |
| Social Mobility | 69.76% | 59.34% |
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 | Gemini 4 Argon cost | GPT-6.1 Sol cost | Gemini 4 Argon latency | GPT-6.1 Sol latency |
|---|---|---|---|---|
| Vals Finance Index | $6.53 | $1.36 | 22m41s | 19m16s |
| Vals Index | $13.37 | $3.61 | 44m58s | 44m24s |
| Vals Legal Index | $6.82 | $2.75 | 25m41s | 31m40s |
| Vals RSI Index | $179.80 | $98.21 | 90h00m | 90h00m |
| Vals Science Index | $18.40 | $1.36 | 57m00s | 33m11s |
| HLAB-Vals-1.1 | $10.55 | $3.54 | 29m15s | 40m26s |
| Legal Research Bench | $6.45 | $3.38 | 31m21s | 34m20s |
| EMB v1.1 | $11.80 | $1.14 | 38m26s | 18m28s |
| Finance Agent v2.1 | $4.32 | $1.63 | 13m13s | 19m05s |
| Tax Agent Bench | $3.47 | $1.32 | 16m26s | 20m14s |
| MedCode | N/A | N/A | 79.53s | 115.34s |
| MedScribe | N/A | N/A | 87.44s | 2m50s |
| ProofBench v1.1 | $0.78 | $0.23 | 6m20s | 4m45s |
| BioMysteryBench | $5.71 | $0.37 | 39m57s | 8m51s |
| MysteryMechanism | $4.59 | $0.28 | 24m55s | 7m17s |
| Robo-bench | $17.33 | $4.97 | 2m32s | 79.89s |
| Terminal-Bench-Science | $44.90 | $3.44 | 1h46m | 1h23m |
| SAGE | N/A | N/A | 63.07s | 50.22s |
| Code Migration v1.1 | $44.79 | $9.51 | 2h22m | 1h39m |
| IOI | $5.31 | $1.26 | 24m19s | 31m31s |
| ProgramBench | $17.66 | $2.87 | 1h11m | 1h10m |
| SWE-bench Multimodal | $11.40 | $0.35 | 33m38s | 6m56s |
| Terminal-Bench 4.0 | $17.64 | $1.72 | 53m35s | 45m12s |
| Vibe Code Bench v1.1 | $8.11 | $6.23 | 37m28s | 1h15m |
| CyberBench v1.1 | $5.11 | $0.29 | 19m10s | 3m42s |
| SRE Bench | $30.45 | $2.69 | 1h35m | 32m32s |
| Public Benefits Bench v1.1 | $2.79 | $0.69 | 26m07s | 50m24s |
Results available only for Gemini 4 Argon
- LegalBench
- Vibe Code Bench 1-100
- CUA-bench
- Time Horizon Index: KSP
Results available only for GPT-6.1 Sol
- ClaimBench