Kimi K3 vs Muse Wonder
Muse Wonder leads Kimi K3 by 6.0 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Kimi K3 wins 3 and Muse Wonder wins 10 of 13 shared benchmarks. Kimi K3 is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | Kimi K3 | Muse Wonder |
|---|---|---|
| Agentic coding | 63 | 68 |
| Long-horizon autonomy | 65 | 70 |
| Knowledge-work agents | 42 | 50 |
| Computer use | 43 | 50 |
| Security & context learning | 53 | 57 |
Shared benchmarks
| Benchmark | Kimi K3 | Muse Wonder |
|---|---|---|
| APEX-Agentsindex | 47.8% | 66.1% |
| Vending-Bench 2index | $5,165 | $9,432 |
| PostTrainBenchindex | 31.5% | 24.2% |
| CritPt | 23.4% | 28.3% |
| DTBench | 91.2% | 95.2% |
| FrontierMath Tier 4 | 39.0% | 70.7% |
| FrontierMath Tiers 1-3 | 72.2% | 82.1% |
| LMCA | 52.7% | 56.8% |
| MirrorCode | 54.4% | 45.6% |
| OTIS Mock AIME | 97.2% | 99.4% |
| ProofBench | 87.0% | 84.0% |
| SciCode | 59.5% | 60.3% |
| WeirdML v3 | 7.5% | 15.1% |
Specs and price
| Kimi K3 | Muse Wonder | |
|---|---|---|
| Agent Index | 54.9 (#26) | 60.9 (#15) |
| Agent $/M | $1.39 | $1.69 |
| Input / output $/M | $3.00 / $15.00 | $4.00 / $20.00 |
| Context | 1M | 1M |
| Released | Jul 16, 2026 | Oct 12, 2026 |
| Weights | Open | Closed |
| API hosts | 29 | 2 |
Don’t pick one. Route.
Shogo can run Kimi K3 and Muse Wonder side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo