$ diff kimi-k3 glm-5-2
Kimi K3 vs GLM-5.2
Moonshot AI and Z.ai, 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
- 1M
- weights
- open
- $/M in
- —
- $/M out
- —
What actually differs
- Context window
- Identical — both accept 1M tokens, so context length is not a reason to pick either.
- Weights
- Both publish weights — Kimi K3 under Modified MIT (commercial use permitted; weights publicly downloadable), GLM-5.2 under MIT (unrestricted, "no regional limits"). Check the licences rather than assuming they permit the same commercial use.
- Recency
- Kimi K3 shipped 33 days after GLM-5.2 (2026-07-16 vs 2026-06-13).
Full spec
| Attribute | Kimi K3 | GLM-5.2 |
|---|---|---|
| Developer | Moonshot AI | Z.ai |
| Released | 2026-07-16 | 2026-06-13 |
| Context window | 1M tokens | 1,000,000 tokens |
| Pricing | $0.30/M cache-hit input · $3.00/M cache-miss input · $15.00/M output | not publicly listed as of 2026-07-13 (~1/6th GPT-5.5 API cost per VentureBeat) |
| License | Modified MIT (commercial use permitted; weights publicly downloadable) | MIT (unrestricted, "no regional limits") |
| Availability | Kimi Code, Kimi app, Kimi API; weights on Hugging Face (`moonshot-ai/kimi-k3`, released 2026-07-27) | GLM Coding Plan (highest subscription tier), standalone API, open weights, NVIDIA NIM hosted |
Values come from Kimi K3 and GLM-5.2, 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.