$ diff kimi-k3 grok-4-5
Kimi K3 vs Grok 4.5
Moonshot AI and xAI, side by side. Every value below is the one recorded on the model’s own wiki page, with its citation — nothing is estimated.
Moonshot AIKimi K3
- context
- 1M
- weights
- open
- $/M in
- $3
- $/M out
- $15
- context
- 500K
- weights
- —
- $/M in
- $2
- $/M out
- $6
What actually differs
- Context window
- Kimi K3 takes 1M against 500K — 2× more room in a single request.
- Input price
- Grok 4.5 at $2/M against $3/M — 1.5× cheaper to feed. Standard rates: one of these labs also quotes a lower cached-input tier, which applies only when a prefix is reused — see the full spec below.
- Output price
- Grok 4.5 at $6/M against $15/M — 2.5× cheaper to generate. Output dominates the bill on most agentic workloads, where the model writes far more than it reads.
- Recency
- Kimi K3 shipped 7 days after Grok 4.5 (2026-07-16 vs 2026-07-09).
Full spec
| Attribute | Kimi K3 | Grok 4.5 |
|---|---|---|
| Developer | Moonshot AI | xAI |
| Released | 2026-07-16 | 2026-07-09 |
| Context window | 1M tokens | 500K tokens |
| Pricing | $0.30/M cache-hit input · $3.00/M cache-miss input · $15.00/M output | $2/M input · $6/M output |
| License | Modified MIT (commercial use permitted; weights publicly downloadable) | not recorded |
| Availability | Kimi Code, Kimi app, Kimi API; weights on Hugging Face (`moonshot-ai/kimi-k3`, released 2026-07-27) | Grok Build (default model), Cursor (all plans), SpaceXAI console/API; not in EU (regulatory review pending as of 2026-07-09) |
Values come from Kimi K3 and Grok 4.5, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing. How these pages are produced.