Models / Compare /Kimi K2.6 vs GLM 5.2
Head to head
Kimi K2.6 vs GLM 5.2
Compared on live catalog data — pricing synced daily, benchmark scores with provenance.
The short version
- GLM 5.2 is cheaper to run hosted: $2.42/M output vs $2.48/M for Kimi K2.6 (1.0× difference at the cheapest tracked provider).
- GLM 5.2 takes 1M tokens of context vs 262K for Kimi K2.6.
- Kimi K2.6 has downloadable weights (OPEN*) so you can self-host it; we hold no downloadable copy of GLM 5.2, so it runs through a provider's API.
- GLM 5.2 scores 73.16 on LiveBench vs 70.54 via Thinking for Kimi K2.6.
- GLM 5.2 scores 51.77 on LiveBench Agentic Coding vs 46.92 via Thinking for Kimi K2.6.
| Spec | ||
|---|---|---|
| Vendor | Moonshot AI | Z.AI |
| Parameters | 1.1T / 32B | 753B |
| 262K | 1M | |
| Weights | OPEN* | OPEN |
| License | Custom licence | MIT License |
| Modality | text+image->text | text->text |
| Released | Apr 14, 2026 | Jun 16, 2026 |
| 667.5 GBest | 475 GBest | |
| Cheapest hosted | $0.59 / $2.48 per 1M (OpenRouter) | $0.76 / $2.42 per 1M (OpenRouter) |
| Tracked providers | 26 | 40 |
Kimi K2.6 — our take
Kimi K2.6 is a large mixture-of-experts model from Moonshot AI that accepts text and images and handles up to 262,144 tokens in a single request. It is the strongest coder in our measured set, with a coding score well above its general chat rating, though its creative writing lags behind both.
full breakdown →GLM 5.2 — our take
Available from more hosts than any other model we track, this large downloadable model from Zhipu carries a permissive MIT licence and a one-million-token request limit. It is a strong choice for teams that want top-tier capability without licensing restrictions.
full breakdown →Need more columns? Use theinteractive comparison to add up to four models.