GPT 5.4 Mini vs MiMo V2.5 Pro: Benchmark Comparison
GPT 5.4 Mini has the higher score on 6 of 14 shared benchmarks; MiMo V2.5 Pro leads on 8.
The largest observed score gap is 13.86 pts on Vibe Code Bench v1.1 , where GPT 5.4 Mini leads.
Reported ±1 standard-error ranges overlap on 7 of 14 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 | MiMo V2.5 Pro | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 33.17% ±1.28 | 33.84% ±1.17 | 0.67 pts | Reported ±1 SE ranges overlap |
| Harvey's Legal Agent Benchmark | 0.00% ±0.00 | 2.08% ±0.83 | 2.08 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 12.50% ±2.30 | 15.87% ±2.54 | 3.37 pts | Reported ±1 SE ranges overlap |
| EMB | 45.43% ±3.65 | 55.23% ±3.11 | 9.80 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 45.36% ±0.45 | 41.50% ±1.19 | 3.86 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 36.59% ±3.10 | 34.79% ±2.62 | 1.80 pts | Reported ±1 SE ranges overlap |
| TaxEval v2 | 71.22% ±0.90 | 73.79% ±0.86 | 2.58 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 83.08% ±2.46 | 82.58% ±2.47 | 0.50 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 84.55% ±0.36 | 84.59% ±0.45 | 0.04 pts | Reported ±1 SE ranges overlap |
| Code Migration | 12.94% ±3.64 | 21.56% ±4.13 | 8.62 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 81.47% ±1.09 | 81.35% ±1.07 | 0.11 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 73.00% ±1.99 | 74.00% ±1.96 | 1.00 pts | Reported ±1 SE ranges 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 | 34.11% ±4.53 | 13.86 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | GPT 5.4 Mini average | MiMo V2.5 Pro average |
|---|---|---|
| Legal | 6.25% | 8.97% |
| Finance | 49.65% | 51.33% |
| Academic | 83.82% | 83.59% |
| Coding | 43.58% | 42.31% |
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 | MiMo V2.5 Pro cost | GPT 5.4 Mini latency | MiMo V2.5 Pro latency |
|---|---|---|---|---|
| Vals Index | $1.45 | $0.16 | 29m49s | 34m33s |
| Harvey's Legal Agent Benchmark | $1.00 | $0.11 | 11m21s | 9m04s |
| Legal Research Bench | $1.86 | $0.08 | 37m36s | 6m56s |
| EMB | $2.15 | $0.22 | 43m21s | 29m08s |
| Finance Agent (v2) | $1.20 | $0.21 | 35m54s | 8m17s |
| Tax Agent Bench | $0.81 | $0.04 | 16m46s | 6m17s |
| TaxEval v2 | N/A | N/A | 44.58s | 62.83s |
| GPQA Diamond | N/A | N/A | 63.11s | 4m11s |
| MMLU Pro | N/A | N/A | 20.00s | 67.76s |
| Code Migration | $1.95 | $0.24 | 28m53s | 48m53s |
| LiveCodeBench | N/A | N/A | 3m09s | 4m31s |
| SWE-bench | $0.51 | $0.02 | 5m26s | 7m51s |
| Terminal-Bench 4.0 | $1.43 | $0.23 | 29m41s | 2h07m |
| Vibe Code Bench v1.1 | $1.19 | $0.17 | 34m15s | 40m04s |
Results available only for GPT 5.4 Mini
- MortgageTax
- MMMU Pro
- SAGE
- ProgramBench
Results available only for MiMo V2.5 Pro
- LegalBench
- MedCode
- MedScribe
- ProofBench v1.1