搞真的啊

[搞真的啊 @PatricCecil]

@morphllm @opperai and even often no response at all

[搞真的啊 @PatricCecil]

@morphllm @opperai model sla is so low, deepseek v4 flash only has 10-20 tokes/s in most of time, as well as glm5.2

[搞真的啊 @PatricCecil]

@morphllm glm cant use tools and thinking format is weird(thinking content appears at reasoning/reasoning content before, but currently appears at messages.content area with </think> block) since yesterday

[搞真的啊 @PatricCecil]

@tejasybhakta glm model cache seems not working? but i saw the cache price at pricing page, will cache support in future?

[搞真的啊 @PatricCecil]

@morphllm how to use this compact in claude code cli?

[搞真的啊 @PatricCecil]

@v0 please, fix https://t.co/E4UIawhiZm endpoint, everything on this endpoint returns 404

[搞真的啊 @PatricCecil]

@Dopamine_Driven found same problem earlier🥲

[搞真的啊 @PatricCecil]

@v0 v0-1.5-md not avaliable on https://t.co/E4UIawhiZm anymore? It returns 404 currently

[搞真的啊 @PatricCecil]

@pablostanley @v0 I mean v0-max and v0-pro are only available on https://t.co/ODKq5AJ54e? I saw in the docs that v0 provides a v1/chat/completions compatible endpoint, which i can use directly with the Codex CLI. But it looks like the max and pro models aren’t available through that endpoint yet

[搞真的啊 @PatricCecil]

@claudeai where is my 1h cache? Today only has 5m cache....

[搞真的啊 @PatricCecil]

@AnthropicAI where is my 1h cache? Today only has 5m cache....