token.app › Compare › Kimi K3 vs Grok 4.6
Kimi K3 vs Grok 4.6
Kimi K3 and Grok 4.6 split the 4 shared benchmarks evenly. Grok 4.6 is 1.1× cheaper per token on a blended 3:1 basis. Kimi K3 accepts the larger context window (1.0M).
Head to head
Blended price uses a 3:1 input:output mix
| Kimi K3 | Grok 4.6 | |
|---|---|---|
| Provider | Moonshot AI | SpaceXAI |
| Input $/1M | $0.239 | $2.00 |
| Output $/1M | $11.94 | $6.00 |
| Blended $/1M | $3.16 | $3.00 |
| Context window | 1.0M | 500K |
| Max output | 944K | 450K |
| Released | Jul 2026 | Aug 2026 |
| Licence | Open weights | Proprietary |
| GPQA Diamond | 93.1% | 94.0% |
| FrontierMath T1–3 | 72.2% | 66.0% |
| FrontierMath T4 | 39.0% | 31.7% |
| OTIS Mock AIME | 97.2% | 99.2% |
Benchmark scores independently run and published by Epoch AI, 'AI Benchmarking Hub', used under CC BY 4.0. Only benchmarks with a score for BOTH models are compared.
Choose Kimi K3 if…
- Higher FrontierMath T1–3 (72.2% vs 66.0%)
- Higher FrontierMath T4 (39.0% vs 31.7%)
- Larger context window (1.0M)
- Open weights — self-hostable
Choose Grok 4.6 if…
- Costs less per token — $3.00 vs $3.16 blended
- Higher GPQA Diamond (94.0% vs 93.1%)
- Higher OTIS Mock AIME (99.2% vs 97.2%)
- Newer model (released Aug 2026)
FAQ
Which is better, Kimi K3 or Grok 4.6?
Across the 4 benchmarks both models have independently-run scores for, Kimi K3 leads on 2 and Grok 4.6 leads on 2. Benchmarks are one input — price and context window are on this page too.
Is Kimi K3 cheaper than Grok 4.6?
Grok 4.6 is cheaper: $3.00 versus $3.16 per million tokens blended at 3:1 input:output. Input alone: $0.239 vs $2.00. Output alone: $11.94 vs $6.00.
What context windows do Kimi K3 and Grok 4.6 have?
Kimi K3: 1,048,576 tokens. Grok 4.6: 500,000 tokens.
Who makes Kimi K3 and Grok 4.6?
Kimi K3 is made by Moonshot AI. Grok 4.6 is made by SpaceXAI.