MJV

[MJV @mjvvjm]

@thsottiaux Give us the ability to use 5.6 Sol in our custom Agents in ChatGPT web.

[MJV @mjvvjm]

Hey @thsottiaux and @OpenAI , when are we getting the ability to set 5.6 Sol on our Agents in ChatGPT web? It only shows 5.5 still.

Please fix!

@ChatGPTapp

[MJV @mjvvjm]

What I love about GPT-5.6 Sol is the follow-through. In Hermes it feels less like a chatbot and more like a dependable operator: it manages ongoing projects, works across my Mac and services, catches mistakes, and keeps going until the job is done. That is why I switched.

[MJV @mjvvjm]

@athyuttamre Awesome, please announce it when the business plans get it as well. Can’t wait!

[MJV @mjvvjm]

@gdb When are business plans getting it?

[MJV @mjvvjm]

@athyuttamre Oh, so it’s only for plus accounts?

[MJV @mjvvjm]

@NousResearch Issue here. When I try to update it after hitting the download button, it shows that it downloads the new version but when it closes it doesn’t do anything after that. Then when I open the app again, it shows the same version as when I originally installed it the day it came out

[MJV @mjvvjm]

@vincent_koc So it works like Hermes by automatically creating new skills?

[MJV @mjvvjm]

@mjgore812 @waglyvibes They are so stupid. The dog is clearly in distress

[MJV @mjvvjm]

@fidexcode 5

[MJV @mjvvjm]

@GeminiApp @BoogieD262 Please fix the audio bug on Gemini Thinking in the web! It is a super useful tool and now it can't process audio files

[MJV @mjvvjm]

@elonmusk @grok The web version can’t handle large files like long PDFs without throwing a “too large” error. Other systems easily manage up to 514 MB of documents (typically up to 10 attachments) without issues.

[MJV @mjvvjm]

@GoogleAI @GoogleDeepMind @OfficialLoganK @GeminiApp You need to fix the audio file processing error Gemini 3 Pro “thinking” has on the web. 2.5 Pro used to transcribe and process audio files accurately without issues. Since Gemini 3 came out, the “thinking” model admits it can’t process files and “simulates” contents based on filenames. The “fast” model accesses and processes audio files fine. Files like .m4a used to work fine. Please fix this as it’s a huge issue for me and probably many others.

[MJV @mjvvjm]

@OfficialLoganK Hey Logan, Gemini 3 Pro “Thinking” cannot process audio files and hallucinates the content. Can you provide an ETA for the fix?

[MJV @mjvvjm]

@OfficialLoganK I would like to use the API key even if I don’t hit the limits to make sure you don’t train on my data.

[MJV @mjvvjm]

@NotebookLM Can you provide support for video files as a source, not just audio? For example, the ability to use a Teams recording so NotebookLM can summarize and organize discussed items.

[MJV @mjvvjm]

@OfficialLoganK @HCSolakoglu Are you guys planning on increasing the rate anytime soon? I’ll subscribe to a ultra when the usage is close to the pro plan on OpenAI

[MJV @mjvvjm]

@iruletheworldmo Code please 😁

[MJV @mjvvjm]

@xai @xai Please enhance Grok’s capabilities to recognize and comprehend diagrams, drawings, and other visual representations. For instance, it should be able to interpret construction drawings effectively.