Kimi K3 vs Mistral Large 4: Benchmark Comparison
Kimi K3 has the higher score on 13 of 16 shared benchmarks; Mistral Large 4 leads on 3.
The largest observed score gap is 77.00 pts on ProofBench v1.1 , where Kimi K3 leads.
Reported ±1 standard-error ranges overlap on 7 of 16 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 | Kimi K3 | Mistral Large 4 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Finance Index | 57.95% ±1.43 | 56.90% ±1.42 | 1.06 pts | Reported ±1 SE ranges overlap |
| Vals Legal Index | 71.72% ±1.80 | 65.61% ±1.63 | 6.11 pts | Reported ±1 SE ranges do not overlap |
| HLAB-Vals-1.1 | 12.92% ±2.81 | 17.50% ±3.22 | 4.58 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 46.15% ±3.46 | 31.73% ±3.23 | 14.42 pts | Reported ±1 SE ranges do not overlap |
| EMB v1.1 | 71.26% ±2.28 | 68.97% ±2.34 | 2.29 pts | Reported ±1 SE ranges overlap |
| Finance Agent v2.1 | 60.01% ±2.09 | 61.38% ±2.02 | 1.37 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 33.37% ±3.39 | 30.39% ±3.31 | 2.98 pts | Reported ±1 SE ranges overlap |
| MedCode | 49.36% ±2.20 | 40.69% ±2.17 | 8.67 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 88.05% ±1.98 | 80.43% ±2.00 | 7.61 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 87.00% ±3.38 | 10.00% ±3.02 | 77.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 72.59% ±1.61 | 67.04% ±2.67 | 5.56 pts | Reported ±1 SE ranges do not overlap |
| IOI | 48.94% ±9.82 | 45.28% ±3.78 | 3.67 pts | Reported ±1 SE ranges overlap |
| SWE-bench Multimodal | 44.17% ±2.27 | 33.13% ±2.15 | 11.04 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 17.17% ±0.51 | 27.78% ±1.82 | 10.61 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench 1-100 | 18.24% ±3.92 | 14.26% ±3.21 | 3.98 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 84.97% ±2.75 | 78.40% ±3.55 | 6.56 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Kimi K3 average | Mistral Large 4 average |
|---|---|---|
| Legal | 29.54% | 24.62% |
| Finance | 54.88% | 53.58% |
| Healthcare | 68.70% | 60.56% |
| Math | 87.00% | 10.00% |
| Science | 72.59% | 67.04% |
| Coding | 42.70% | 39.77% |
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 | Kimi K3 cost | Mistral Large 4 cost | Kimi K3 latency | Mistral Large 4 latency |
|---|---|---|---|---|
| Vals Finance Index | $2.58 | $2.86 | 22m24s | 51m50s |
| Vals Legal Index | $3.47 | $4.08 | 31m51s | 1h09m |
| HLAB-Vals-1.1 | $4.14 | $4.18 | 42m36s | 58m04s |
| Legal Research Bench | $3.47 | $5.33 | 12m40s | 1h33m |
| EMB v1.1 | $3.07 | $4.84 | 12m21s | 1h22m |
| Finance Agent v2.1 | $1.88 | $1.02 | 14m35s | 16m26s |
| Tax Agent Bench | $2.79 | $2.72 | 40m15s | 57m18s |
| MedCode | N/A | N/A | 36.99s | 68.68s |
| MedScribe | N/A | N/A | 38.24s | 2m05s |
| ProofBench v1.1 | $1.65 | $0.64 | 34m31s | 20m59s |
| BioMysteryBench | $1.06 | $2.07 | 4m55s | 25m32s |
| IOI | $14.17 | $5.39 | 4h18m | 1h24m |
| SWE-bench Multimodal | $1.65 | $5.14 | 22m45s | 27m27s |
| Terminal-Bench 4.0 | $12.48 | $10.25 | 3h09m | 1h40m |
| Vibe Code Bench 1-100 | $9.40 | $6.75 | 1h48m | 59m53s |
| Vibe Code Bench v1.1 | $10.01 | $39.46 | 16m39s | 2h53m |
Results available only for Kimi K3
- Vals Index
- Vals Inference Provider Index
- Vals RSI Index
- LegalBench
- MortgageTax
- TaxEval v2
- Terminal-Bench-Science
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- SAGE
- Code Migration v1.1
- LiveCodeBench
- ProgramBench
- SWE-bench
- CyberBench v1.1
- Public Benefits Bench v1.1
- Time Horizon Index: KSP
Results available only for Mistral Large 4
- MysteryMechanism