Owen · Zmemo

Learning by testing ideas into workflows.

[Owen · Zmemo @theZmemo]

@goan999999 好方法收藏

[Owen · Zmemo @theZmemo]

@getsome_air @getsome_air
Gemini CLI’s official ACP support is gone after Google shut it down for consumers.
Please add support for Antigravity (Google’s new CLI) instead. Thanks!

[Owen · Zmemo @theZmemo]

ACP agents are set up and running tasks fine. Adding context via @file also works.
But selecting code lines in the editor and clicking “Add to Task” still forces a login prompt (it says it’s not a subscription issue).
Active task + ACP agent selected, still blocked.
Looks like the editor selection shortcut isn’t fully adapted for pure ACP setups yet. Would appreciate a fix, thanks!

[Owen · Zmemo @theZmemo]

@0x_Northland 我的也没恢复 instant正常 其他还是5.5mini

[Owen · Zmemo @theZmemo]

@shengtang135754 我这instant正常的，但切换 medium和high就变成5.5mini ，选5.5和5.6都是非instant同样的到5.5mini

[Owen · Zmemo @theZmemo]

@goan999999 这个好像现在不现实具体用的哪个模型了 只有消耗额度

[Owen · Zmemo @theZmemo]

@caishengye01 @GKev1n 我也是 只有instant是5.6 切其他都是5.5mini

[Owen · Zmemo @theZmemo]

@bil0090 Doesn't work

[Owen · Zmemo @theZmemo]

@thsottiaux Just renewed annual membership on iOS.
Instant correctly runs 5.6, but switching to Medium/High drops to 5.5-mini with zero thinking process — same on web, Desktop and iOS app.
Silent downgrade for paying users. Any explanation?

[Owen · Zmemo @theZmemo]

Nah, it’s clearly not a misunderstanding.
When I switch to Medium / High, it instantly spits out the answer with zero thinking process.
But on Instant, it actually shows the reasoning steps.
A real higher-tier model should think more and respond slower.
Now the higher settings are faster with no thinking — that’s literally being routed to a lighter model.

[Owen · Zmemo @theZmemo]

Just paid for the annual sub, and this is what I get:
Selecting Medium or High makes replies faster and quality drops straight to 5.5-mini.
UI shows “5.6 Sol Medium/High”, but it’s clearly being silently routed to a smaller model.
Is this intentional downgrade or a routing bug?
Would love a clear explanation.@thsottiaux @sama #ChatGPT #OpenAI #GPT

[Owen · Zmemo @theZmemo]

@thsottiaux @thsottiaux

[Owen · Zmemo @theZmemo]

@thsottiaux Just got the annual plan, but selecting 5.6 sol Medium/High silently routes to 5.5-mini — any explanation?
https://t.co/GV2OBSElcR

[Owen · Zmemo @theZmemo]

Just paid for the annual sub, and this is what I get:
Selecting Medium or High makes replies faster and quality drops straight to 5.5-mini.
UI shows “5.6 Sol Medium/High”, but it’s clearly being silently routed to a smaller model.
Is this intentional downgrade or a routing bug?
Would love a clear explanation.@thsottiaux @sama #ChatGPT #OpenAI #GPT

[Owen · Zmemo @theZmemo]

@arkuy99 刚买了gpt一年的🤣

[Owen · Zmemo @theZmemo]

@elonmusk Usage really does drain super fast, so more free quota is perfectly timed. Keep Grok going strong 🚀

[Owen · Zmemo @theZmemo]

@YinsenW_ 这个好👍🏻

[Owen · Zmemo @theZmemo]

@GeekCatX 差太多了吧

[Owen · Zmemo @theZmemo]

@xmgplus @tibo 昨晚刚刚自动重置还没用😂

[Owen · Zmemo @theZmemo]

@thsottiaux A lot of people are getting their natural reset today or tomorrow (mine’s in ~10 hours) — any chance of a free reset card? 👀