Wonder Full vs GPT-5.6 Sol: Benchmark Comparison
Wonder Full has the higher score on 15 of 19 shared benchmarks; GPT-5.6 Sol leads on 4.
The largest observed score gap is 20.20 pts on Terminal-Bench 4.0 , where Wonder Full leads.
Reported ±1 standard-error ranges overlap on 6 of 18 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 | Wonder Full | GPT-5.6 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 24.43% | 23.88% | 0.55 pts | Uncertainty comparison unavailable |
| Vals Science Index | 48.54% ±2.23 | 41.48% ±1.93 | 7.06 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 53.37% ±3.47 | 48.08% ±3.47 | 5.29 pts | Reported ±1 SE ranges overlap |
| LegalBench | 87.42% ±0.41 | 86.97% ±0.41 | 0.45 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 43.00% ±3.57 | 31.09% ±3.34 | 11.91 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 49.93% ±2.18 | 43.97% ±2.26 | 5.96 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 89.33% ±2.00 | 85.23% ±1.97 | 4.10 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 94.00% ±2.39 | 83.00% ±3.77 | 11.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 73.33% ±0.64 | 71.11% ±0.64 | 2.22 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 35.13% ±3.21 | 33.33% ±3.17 | 1.80 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench-Science | 37.14% ±5.82 | 20.00% ±4.82 | 17.14 pts | Reported ±1 SE ranges do not overlap |
| IOI | 79.83% ±2.99 | 91.17% ±4.51 | 11.33 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.50% ±0.50 | 1.50% ±0.86 | 1.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 58.08% ±3.31 | 37.88% ±0.00 | 20.20 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench 1-100 | 33.98% ±5.32 | 20.01% ±3.63 | 13.97 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 86.00% ±1.93 | 80.50% ±3.72 | 5.50 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 73.63% ±5.56 | 76.31% ±5.18 | 2.68 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 11.83% ±2.00 | 30.53% ±2.85 | 18.70 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 73.75% ±1.15 | 66.51% ±1.23 | 7.24 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Wonder Full average | GPT-5.6 Sol average |
|---|---|---|
| Legal | 70.39% | 67.52% |
| Finance | 43.00% | 31.09% |
| Healthcare | 69.63% | 64.60% |
| Math | 94.00% | 83.00% |
| Science | 48.54% | 41.48% |
| Coding | 51.68% | 46.21% |
| Cyber | 42.73% | 53.42% |
| Social Mobility | 73.75% | 66.51% |
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 | Wonder Full cost | GPT-5.6 Sol cost | Wonder Full latency | GPT-5.6 Sol latency |
|---|---|---|---|---|
| Vals RSI Index | N/A | N/A | 90h00m | 90h00m |
| Vals Science Index | $10.15 | $3.40 | 1h34m | 43m32s |
| Legal Research Bench | $3.60 | $21.61 | 27m34s | 1h17m |
| LegalBench | N/A | N/A | 38.27s | 6.20s |
| Tax Agent Bench | $2.15 | $7.51 | 18m04s | 44m41s |
| MedCode | N/A | N/A | 3m03s | 96.59s |
| MedScribe | N/A | N/A | 115.72s | 94.40s |
| ProofBench v1.1 | $0.86 | $1.44 | 11m36s | 8m40s |
| BioMysteryBench | $3.47 | $1.93 | 26m14s | 18m18s |
| MysteryMechanism | $4.16 | $0.96 | 41m34s | 8m43s |
| Terminal-Bench-Science | $22.81 | $7.32 | 3h35m | 1h44m |
| IOI | $7.08 | $7.98 | 1h31m | 1h02m |
| ProgramBench | $28.65 | $15.29 | 6h04m | 29m24s |
| Terminal-Bench 4.0 | $10.53 | $7.98 | 1h12m | 33m46s |
| Vibe Code Bench 1-100 | $19.54 | $47.43 | 1h57m | 2h37m |
| Vibe Code Bench v1.1 | $2.29 | $33.40 | 10m48s | 33m53s |
| CyberBench v1.1 | $3.63 | $3.05 | 35m27s | 10m47s |
| SRE Bench | $33.81 | $42.52 | 5h34m | 1h28m |
| Public Benefits Bench v1.1 | $4.64 | $8.85 | 1h13m | 1h01m |
Results available only for Wonder Full
- 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
Results available only for GPT-5.6 Sol
- MortgageTax
- TaxEval v2
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- SAGE
- LiveCodeBench
- SkillsBench
- SWE-bench
- CUA-bench
- Time Horizon Index: KSP