GPT-6 Sol vs Kimi K3
GPT-6 Sol leads Kimi K3 by 8.7 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. GPT-6 Sol wins 17 and Kimi K3 wins 3 of 20 shared benchmarks. GPT-6 Sol is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | GPT-6 Sol | Kimi K3 |
|---|---|---|
| Agentic coding | 72 | 63 |
| Long-horizon autonomy | 76 | 65 |
| Knowledge-work agents | 48 | 42 |
| Computer use | 54 | 43 |
| Security & context learning | 59 | 53 |
Shared benchmarks
| Benchmark | GPT-6 Sol | Kimi K3 |
|---|---|---|
| APEX-Agentsindex | 54.3% | 47.8% |
| Vending-Bench 2index | $14,428 | $5,165 |
| ARC-AGI | 95.5% | 94.5% |
| ARC-AGI-2 | 89.6% | 60.4% |
| Chess Puzzles | 56.0% | 37.0% |
| CritPt | 30.9% | 23.4% |
| DTBench | 97.3% | 91.2% |
| FrontierCode | 49.3% | 44.2% |
| FrontierMath Tier 4 | 90.0% | 39.0% |
| FrontierMath Tiers 1-3 | 89.8% | 72.2% |
| GPQA Diamond | 94.3% | 93.1% |
| LMCA | 59.1% | 52.7% |
| MirrorCode | 48.9% | 54.4% |
| OTIS Mock AIME | 100.0% | 97.2% |
| ProofBench | 83.0% | 87.0% |
| SciCode | 57.6% | 59.5% |
| SimpleBench | 73.1% | 60.7% |
| SimpleQA Verified | 60.7% | 50.6% |
| WebDev Arena | 1689 | 1674 |
| WeirdML v3 | 19.7% | 7.5% |
Specs and price
| GPT-6 Sol | Kimi K3 | |
|---|---|---|
| Agent Index | 63.6 (#13) | 54.9 (#26) |
| Agent $/M | $0.92 | $1.39 |
| Input / output $/M | $2.00 / $10.00 | $3.00 / $15.00 |
| Context | 1.1M | 1M |
| Released | Sep 22, 2026 | Jul 16, 2026 |
| Weights | Closed | Open |
| API hosts | 5 | 29 |
Don’t pick one. Route.
Shogo can run GPT-6 Sol and Kimi K3 side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo