Body:
Hi OpenAI team and fellow developers,
I’m a long-time heavy user of ChatGPT, particularly using it in Canvas (project) mode for real, productivity-grade creative work — including writing technical specs, drafting patent applications, building pitch decks, and outlining innovation models.
Issue
Currently, when working within the Canvas / Project interface, the voice input feature becomes completely unavailable, even though it functions normally in standard chat mode. This behavior causes a significant workflow disruption, especially for those of us relying on ChatGPT as a cross-modal cognitive partner (text + voice + images).
Consequences for real users:
Productivity drops due to forced context-switching (creating new chats just to use voice);
Breaks interaction flow in deep work states;
Prevents multimodal creativity and spontaneous ideation in structured documents;
Feels like an inconsistent experience — users think voice is broken.
Usage Stats (from my account):
Metric Value
Total ChatGPT pages generated 4,783 pages (as of April 2025)
Messages in the last 30 days ~2,600
Days active (last 7 days) 8 days (yes, every day!)
My role Inventor with 30+ years of product R&D
Use case Innovation documentation, technical drafting, AI co-creation
Dependency on voice Daily – for drafting, commands, idea dumping
Suggestions
Please enable voice input in Canvas mode for users who want it;
If not immediately possible, show a clear UI message explaining why it’s unavailable;
Long term: allow “global voice input” that can interact across modes, even inside documents;
Eventually: allow voice → text dictation to insert content directly into Canvas documents.
Why this matters
This request doesn’t come from a casual user — it comes from someone building real things with GPT.
Fixing this inconsistency will hugely improve the experience for advanced users and make GPT more than just a chat tool — a true assistant.
Thank you for building this amazing platform. Looking forward to seeing this evolve.
Best,
(A dedicated GPT user)