JJAYCHEN

Shanghai / ECNU / Currently Software Engineer @NVIDIA, Previously @ARM, @MorganStanley, @TripComGroup, @Tencent

[JJAYCHEN @jjaychen1e]

@arkuy99 seedex 的体验特别好.. 能有 Codex 的话就太好了！

[JJAYCHEN @jjaychen1e]

@arkuy99 企业账号会被限制官方 remote control...

[JJAYCHEN @jjaychen1e]

@arkuy99 请问有计划吗～

[JJAYCHEN @jjaychen1e]

@jpschroeder it seems the implementation does not look as good as the UI, it just contacts everything in plain text, no Cache, no Tools

[JJAYCHEN @jjaychen1e]

@charlieholtz GREAT NEWS! But it seems the MCPs installed in Cursorare not available in Conductor, do I miss some setup?

[JJAYCHEN @jjaychen1e]

@peduarte @das_rdsm @kalyandechiraju @raycast Do we remove the custom provider in v2 intentionally or is it still under development? It's a super important feature for me and I'd like to see it in v2!

[JJAYCHEN @jjaychen1e]

@yetone 为什么 Opus 一直报错 🤔 [Error: Connect error resource_exhausted: Error]

[JJAYCHEN @jjaychen1e]

@calicastle 请问可以参加内测吗！

[JJAYCHEN @jjaychen1e]

@charlieholtz @conductor_build It's nice to have the 1M model!! And could we also provide an option to enable the specific max effort level for Opus 4.6?

[JJAYCHEN @jjaychen1e]

@conductor_build @pierrecomputer Could we also add Sonnet / Opus 4.6 1M models?

[JJAYCHEN @jjaychen1e]

@conductor_build I can't find Opus 4.6 1M, is it available through Conductor?

[JJAYCHEN @jjaychen1e]

@CitronaTeo 多谢！

[JJAYCHEN @jjaychen1e]

小宇宙真是让人又爱又恨，最近发现小宇宙已经支持了实时字幕了。

[JJAYCHEN @jjaychen1e]

@ericzakariasson can GPT and Grok share cache when swtich between them in Cursor?

[JJAYCHEN @jjaychen1e]

@cursor_ai the 1M Claude model just exploded my context, missed all cache, and ate 120$ in one query until I observed the slowness of the process and manually stoped it.

I suspect this is a severe bug of Cursor when using the 1M Claude model

[JJAYCHEN @jjaychen1e]

@MagicableGroot @hi_caicai Reflix 要更新了吗

[JJAYCHEN @jjaychen1e]

@thomaspaulmann @raycast @meetgranola Thank you, it works now! And I have a question that how does Raycast parse the reasoning content? The reasoning content inside <think> ... </think> is displayed as plain text. I'm using an openai-compatable provider.

[JJAYCHEN @jjaychen1e]

@pzx553 @YuTengjing 请问如何配置？

[JJAYCHEN @jjaychen1e]

@thomaspaulmann @raycast @meetgranola yes, I think this is a bug. I get this popup when I switch to models from custom prividers. I run out of the 50 free trial messages, that could be the cause.