GPT 5.4 Mini vs Inkling: Benchmark Comparison
GPT 5.4 Mini has the higher score on 3 of 12 shared benchmarks; Inkling leads on 8.
The largest observed score gap is 28.76 pts on Vibe Code Bench v1.1 , where GPT 5.4 Mini leads.
Reported ±1 standard-error ranges overlap on 4 of 12 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 | GPT 5.4 Mini | Inkling | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 12.50% ±2.30 | 28.36% ±3.13 | 15.86 pts | Reported ±1 SE ranges do not overlap |
| MortgageTax | 63.51% ±0.91 | 63.99% ±0.94 | 0.48 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 8.81% ±2.04 | 15.29% ±2.42 | 6.48 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 71.22% ±0.90 | 75.31% ±0.84 | 4.09 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 83.08% ±2.46 | 87.12% ±1.78 | 4.04 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 84.55% ±0.36 | 86.30% ±0.34 | 1.74 pts | Reported ±1 SE ranges do not overlap |
| MMMU Pro | 79.25% ±0.97 | 78.50% ±0.99 | 0.75 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 81.47% ±1.09 | 85.52% ±1.00 | 4.05 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 |
| SWE-bench | 73.00% ±1.99 | 77.60% ±1.87 | 4.60 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 2.52% ±1.01 | 0.51% ±0.51 | 2.02 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 47.97% ±5.61 | 19.21% ±3.66 | 28.76 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | GPT 5.4 Mini average | Inkling average |
|---|---|---|
| Legal | 12.50% | 28.36% |
| Finance | 47.85% | 51.53% |
| Academic | 82.29% | 83.97% |
| Coding | 40.99% | 36.57% |
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 | GPT 5.4 Mini cost | Inkling cost | GPT 5.4 Mini latency | Inkling latency |
|---|---|---|---|---|
| Legal Research Bench | $1.86 | $0.49 | 37m36s | 14m29s |
| MortgageTax | N/A | N/A | 2m06s | 69.11s |
| Tax Agent Bench | $0.81 | $0.35 | 16m46s | 13m31s |
| TaxEval v2 | N/A | N/A | 44.58s | 3m18s |
| GPQA Diamond | N/A | N/A | 63.11s | 5m09s |
| MMLU Pro | N/A | N/A | 20.00s | 73.26s |
| MMMU Pro | N/A | N/A | 50.81s | 2m04s |
| LiveCodeBench | N/A | N/A | 3m09s | 4m18s |
| ProgramBench | N/A | N/A | 53m45s | 45m34s |
| SWE-bench | $0.51 | $0.59 | 5m26s | 6m34s |
| Terminal-Bench 4.0 | $1.44 | $2.50 | 29m30s | 22m59s |
| Vibe Code Bench v1.1 | $1.19 | $1.61 | 34m15s | 18m25s |
Results available only for GPT 5.4 Mini
None.
Results available only for Inkling
- Vals Index
- Vals Finance Index
- Vals Legal Index
- HLAB-Vals-1.1
- LegalBench
- EMB v1.1
- Finance Agent v2.1
- MedCode
- MedScribe
- ProofBench v1.1
- Code Migration v1.1
- IOI
- SkillsBench
- CyberBench v1.1
- SRE Bench
- Public Benefits Bench v1.1