Grok 4.6 vs Kimi K3: Benchmark Comparison
Grok 4.6 has the higher score on 10 of 24 shared benchmarks; Kimi K3 leads on 13.
The largest observed score gap is 36.00 pts on ProofBench v1.1 , where Kimi K3 leads.
Reported ±1 standard-error ranges overlap on 14 of 23 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 | Grok 4.6 | Kimi K3 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 25.46% | 20.88% | 4.58 pts | Uncertainty comparison unavailable |
| Immigration | 65.98% ±1.53 | 68.26% ±1.51 | 2.27 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 48.08% ±3.47 | 46.15% ±3.46 | 1.92 pts | Reported ±1 SE ranges overlap |
| LegalBench | 86.31% ±0.42 | 86.21% ±0.41 | 0.09 pts | Reported ±1 SE ranges overlap |
| MortgageTax | 64.19% ±0.95 | 66.34% ±0.91 | 2.15 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 33.66% ±3.41 | 33.37% ±3.39 | 0.29 pts | Reported ±1 SE ranges overlap |
| TaxEval v2 | 71.10% ±0.90 | 75.72% ±0.84 | 4.62 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 44.71% ±2.26 | 49.36% ±2.20 | 4.64 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 86.53% ±1.96 | 88.05% ±1.98 | 1.51 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 51.00% ±5.02 | 87.00% ±3.38 | 36.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 72.22% ±0.00 | 72.59% ±1.61 | 0.37 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench-Science | 11.43% ±3.83 | 1.43% ±1.43 | 10.00 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 94.70% ±1.13 | 92.93% ±1.31 | 1.77 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 89.40% ±0.30 | 87.97% ±0.32 | 1.43 pts | Reported ±1 SE ranges do not overlap |
| IOI | 47.61% ±2.09 | 48.94% ±9.82 | 1.33 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 88.22% ±0.94 | 87.19% ±0.97 | 1.04 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 1.00% ±0.70 | 2.00% ±0.99 | 1.00 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 95.60% ±0.92 | 93.40% ±1.11 | 2.20 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 17.17% ±1.34 | 17.17% ±0.51 | 0.00 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 76.24% ±3.82 | 84.97% ±2.75 | 8.73 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 67.68% ±5.87 | 75.24% ±5.56 | 7.56 pts | Reported ±1 SE ranges overlap |
| Access to Justice Benchmark | 71.34% ±0.73 | 68.17% ±0.75 | 3.17 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 66.85% ±1.23 | 68.20% ±1.21 | 1.35 pts | Reported ±1 SE ranges overlap |
| Time Horizon Index: KSP | 5.83% ±0.00 | 10.50% ±5.38 | 4.67 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Grok 4.6 average | Kimi K3 average |
|---|---|---|
| Legal | 66.79% | 66.87% |
| Finance | 56.31% | 58.47% |
| Healthcare | 65.62% | 68.70% |
| Math | 51.00% | 87.00% |
| Science | 41.83% | 37.01% |
| Academic | 92.05% | 90.45% |
| Coding | 54.31% | 55.61% |
| Cyber | 67.68% | 75.24% |
| Social Mobility | 69.09% | 68.19% |
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 | Grok 4.6 cost | Kimi K3 cost | Grok 4.6 latency | Kimi K3 latency |
|---|---|---|---|---|
| Vals RSI Index | $88.12 | $136.82 | 90h00m | 90h00m |
| Immigration | $0.10 | $0.18 | 3m40s | 64.84s |
| Legal Research Bench | $1.53 | $3.47 | 25m04s | 12m40s |
| LegalBench | N/A | N/A | 27.42s | 7.11s |
| MortgageTax | N/A | N/A | 8.55s | 78.94s |
| Tax Agent Bench | $0.98 | $2.79 | 10m58s | 40m15s |
| TaxEval v2 | N/A | N/A | 47.57s | 2m25s |
| MedCode | N/A | N/A | 2m07s | 36.99s |
| MedScribe | N/A | N/A | 70.94s | 38.24s |
| ProofBench v1.1 | $0.82 | $1.65 | 14m22s | 34m31s |
| BioMysteryBench | $1.23 | $1.06 | 14m18s | 4m55s |
| Terminal-Bench-Science | $4.82 | $21.30 | 1h18m | 4h49m |
| GPQA Diamond | N/A | N/A | 3m29s | 2m02s |
| MMLU Pro | N/A | N/A | 54.98s | 38.72s |
| IOI | $8.18 | $14.17 | 3h57m | 4h18m |
| LiveCodeBench | N/A | N/A | 118.40s | 3m20s |
| ProgramBench | $16.76 | $70.48 | 4h27m | 5h42m |
| SWE-bench | $0.78 | $0.76 | 10m04s | 10m19s |
| Terminal-Bench 4.0 | $5.06 | $12.48 | 29m25s | 3h09m |
| Vibe Code Bench v1.1 | $4.88 | $10.01 | 25m30s | 16m39s |
| CyberBench v1.1 | $4.07 | $2.13 | 36m52s | 30m44s |
| Access to Justice Benchmark | $0.71 | $0.45 | 6m13s | 9m42s |
| Public Benefits Bench v1.1 | $0.92 | $0.91 | 25m53s | 7m19s |
| Time Horizon Index: KSP | $225.12 | $216.43 | N/A | N/A |
Results available only for Grok 4.6
- Vals Science Index
- MysteryMechanism
- SkillsBench
Results available only for Kimi K3
- Vals Index
- Vals Finance Index
- Vals Legal Index
- Vals Inference Provider Index
- HLAB-Vals-1.1
- EMB v1.1
- Finance Agent v2.1
- MMMU Pro
- SAGE
- Code Migration v1.1
- SWE-bench Multimodal
- CyberOffense
- Financial Coaching Bench