glm-5.2 launched on June 16 — fully MIT open weights with a 1M context window, topping multiple long-horizon coding leaderboards as the #1 open-source model. APIYI added it the day before yesterday (June 18); it has good word-of-mouth and you can call, test, and use it directly now.
Calling and pricing (FYI):
- Pricing: single top-ups over $100 get 9% off and up, with lower rates for large accounts.
- Format: only the OpenAI-native format is supported for now.
- ClaudeCode: not yet supported (it needs Anthropic’s native protocol endpoint); we’ll announce promptly via
/liveonce it is.
← Back to Live Updates · 📚 Monthly Archive