GPT 5.4 Mini vs Inkling Small: Benchmark Comparison
GPT 5.4 Mini has the higher score on 4 of 12 shared benchmarks; Inkling Small leads on 8.
The largest observed score gap is 28.92 pts on Vibe Code Bench v1.1 , where GPT 5.4 Mini leads.
Reported ±1 standard-error ranges overlap on 3 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 Small | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 12.50% ±2.30 | 25.48% ±3.03 | 12.98 pts | Reported ±1 SE ranges do not overlap |
| MortgageTax | 63.51% ±0.91 | 62.28% ±0.95 | 1.23 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 8.81% ±2.04 | 15.39% ±2.37 | 6.58 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 71.22% ±0.90 | 75.51% ±0.84 | 4.29 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 83.08% ±2.46 | 83.59% ±2.04 | 0.51 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 84.55% ±0.36 | 85.57% ±0.35 | 1.01 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 50.81% ±3.40 | 32.74% ±3.26 | 18.07 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 81.47% ±1.09 | 85.93% ±0.99 | 4.47 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.00% ±0.00 | 0.50% ±0.50 | 0.50 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 73.00% ±1.99 | 82.20% ±1.71 | 9.20 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 2.52% ±1.01 | 0.00% ±0.00 | 2.52 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 47.97% ±5.61 | 19.05% ±3.40 | 28.92 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | GPT 5.4 Mini average | Inkling Small average |
|---|---|---|
| Legal | 12.50% | 25.48% |
| Finance | 47.85% | 51.06% |
| Academic | 83.82% | 84.58% |
| Education | 50.81% | 32.74% |
| Coding | 40.99% | 37.54% |
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 Small cost | GPT 5.4 Mini latency | Inkling Small latency |
|---|---|---|---|---|
| Legal Research Bench | $1.86 | $0.14 | 37m36s | 12m55s |
| MortgageTax | N/A | N/A | 2m06s | 112.21s |
| Tax Agent Bench | $0.81 | $0.09 | 16m46s | 6m25s |
| TaxEval v2 | N/A | N/A | 44.58s | 4m02s |
| GPQA Diamond | N/A | N/A | 63.11s | 4m20s |
| MMLU Pro | N/A | N/A | 20.00s | 71.69s |
| SAGE | N/A | N/A | 2m21s | 3m26s |
| LiveCodeBench | N/A | N/A | 3m09s | 5m02s |
| ProgramBench | N/A | N/A | 53m45s | 28m31s |
| SWE-bench | $0.51 | $0.13 | 5m26s | 4m05s |
| Terminal-Bench 4.0 | $1.44 | $0.32 | 29m30s | 11m45s |
| Vibe Code Bench v1.1 | $1.19 | $0.43 | 34m15s | 49m09s |
Results available only for GPT 5.4 Mini
- MMMU Pro
Results available only for Inkling Small
- Vals Finance Index
- Vals 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
- Public Benefits Bench v1.1