[Feature]: Support direct OpenAI Decisions API routing

Open
  • Is something wrong?
#8,190 2 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
55/100
Issue type
Feature
Clarity
Mostly clear
Activity status
Active
Tech stack
typescript
Domain
backend-api-design

Research direction

Start with src/combos/jev-decision-contract.ts and the Vercel Decisions support in PR #7574; trace how decision provider connection testing and routing select request and response formats. Compare that path with OpenAI’s Decisions API requirements in the linked docs. Done means an OpenAI key limited to Decisions → Write can be tested and used for routing while existing model and effort restrictions, timeouts, fallbacks, and other backends continue to work.

Written by the indexing model from the issue text.

Description

enhancement proxy
Area

Proxy and routing

What are you trying to accomplish?

I'd like to use my OpenAI API credits to choose the model and reasoning effort for a JEV Combo. The selected models would still run through my existing provider connections, ncluding my existing Codex-login provider.

What prevents this today?

The v2.97.0 code only recognises ai-gateway.vercel.sh as a Decisions endpoint. Pointing a decision provider at OpenAI still uses the System One request format.

This comes from checking the source. I haven't tested it with a live OpenAI request.

What should OpenCodex do?

Let users connect directly to OpenAI's Decisions API with their own API key, using OpenAI's request and response format.

Connection testing and routing should work with a key limited to Decisions → Write. The existing model and effort restrictions, timeouts, fallbacks, and other decision backends should keep working as they do today.

Example usage or interface

Add a decision provider with these settings from OpenAI's API docs:

Endpoint: https://api.openai.com/v1/decisions
Model: gpt-6-luna
Authentication: OpenAI API key

Then select it as the decision provider for a JEV Combo containing my existing Codex models.

Alternatives or workarounds

I've considered TypeSafe JEV, Vercel AI Gateway, and using a Responses model as the router. I'd prefer to use the Decisions API directly through OpenAI.

Additional context

PR #7574 added Vercel Decisions support. OpenAI also lists Decisions as covered by eligible ChatGPT API credits.

Checks
  • I searched existing issues and documentation.
  • This request describes a concrete OpenCodex workflow rather than merely naming a desired technology.
  • I removed secrets and personal data.
Dominant language
TypeScript
Stars
16.9k
Forks
1.3k
Avg merge
19h 24m
Merged PRs (30d)
676

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from lidge-jun/opencodex

All issues in lidge-jun/opencodex

Similar issues

More TypeScript issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.