How Do You Use More AI Execution Options Without Signing In Twice?
When you add another way to run an AI agent, you should get more choice—not another installation, another sign-in, or another credential to look after.
Sign in to ChatGPT once, then keep using that same account when you switch execution options. You get more ways to run agents without adding a second authentication workflow to your project.
Add another execution option without adding another setup step.
Pi is built into Finchstory, so you do not need to install a separate CLI. It loads only when you actually run a Pi role. DeepSeek and ChatGPT are supported first, while thinking, tools, attachments, stopping, steering, and retrying stay in the same conversation experience you already use.
One ChatGPT sign-in is enough.
Keep using the existing Codex sign-in. When you run a Pi role with ChatGPT, it uses the same account instead of asking you to sign in again. Roles using that same account also reflect the same available ChatGPT capacity.
Switch execution options without managing another credential.
Your ChatGPT credentials stay out of project files and role settings. Changing how an agent runs does not give you another authentication file or another sensitive secret to maintain inside the project.
Refresh only the model you actually want to check.
In model settings, you see OpenAI GPT and its sign-in state without needing to understand the internal execution path. If one connected remote model needs a refresh, you can refresh that model alone without touching the others or sending a paid inference test.
Finchstory