Project-level persistent memory & Git repository sync across threads
Allow linking a GitHub/GitLab branch directly to a ChatGPT Project workspace with automatic diff indexing so we don't have to re-upload modified files on every commit.
AI & Cloud Product Roadmap · 7 Active Feature Proposals · 30,631 Priority Votes
OpenAI is a leading AI & Cloud platform tracked on CrowdFeatures. Engineers, product managers, and power users use this board to propose new OpenAI capabilities, upvote high-priority roadmap items, and verify features that have already launched in production.
Allow linking a GitHub/GitLab branch directly to a ChatGPT Project workspace with automatic diff indexing so we don't have to re-upload modified files on every commit.
Current fine-tuning only supports supervised learning. Allow customers to use RLHF with custom reward models to truly align fine-tuned models with specific business requirements.
ChatGPT memory is inconsistent and limited. Allow users to build a comprehensive persistent knowledge base that the model reliably references in every conversation.
Many enterprises cannot send data to OpenAI servers due to compliance. Offer a fully self-hosted deployment option for ChatGPT that runs entirely within corporate infrastructure.
The current usage dashboard is too basic. Show real-time token costs per request, daily trending, cost projections, and budget alerts with granular per-model breakdowns.
Model updates silently change behavior and break production applications. Allow pinning to a specific model snapshot version with guaranteed availability for at least 12 months.
Code Interpreter is only available in ChatGPT UI. Expose a sandboxed code execution environment through the API so developers can build applications with safe code running capabilities.