Resources // LLM Tokens
Codex open source fund
- OSS-only
The offer
Part of OpenAI's "Codex for Open Source" program (the fund was folded into a broader program ~March 2026). Selected maintainers get:
- 6 months of ChatGPT Pro (includes Codex)
- Conditional access to Codex Security (GPT-5.x code-security review), granted case-by-case for high-impact/high-risk repos
- API credits from the $1 million Codex Open Source Fund: up to $25,000 per project, reviewed on a rolling basis. Tool-agnostic (use Codex, OpenCode, Cline and others).
Who qualifies
- Be a primary/core maintainer (write access) of a widely-used public GitHub repo
- Valid ChatGPT account; provide GitHub username, repo URL, and OpenAI Organization ID, and justify the project's ecosystem importance
- Apply via the form (you can also nominate another maintainer); awarded on a rolling basis at OpenAI's discretion
Community Insights
This is a maintainer grant for open-source work, not a general credits program for startups. OpenAI’s $1M Codex Open Source Fund gives up to $25k in API credits per project (note: $1M is the total fund, not a per-project amount), and as of ~March 2026 it sits inside the broader “Codex for Open Source” program that leads with 6 months of ChatGPT Pro plus conditional Codex Security access. It is aimed at core maintainers of widely-used public repos, awarded on a rolling basis. Independent discussion is thin (it is a niche grant); the credible signal is that OpenAI has named real recipients.
Best Practices (from community tips)
- Apply as a core maintainer of a genuinely widely-used public repo. Quantify impact (stars, downloads, dependents).
- Use the credits for maintainer automation (Codex-in-PR-review, release workflows). The program is tool-agnostic, so you are not locked to one CLI.
- Treat “$1M” as the fund, “$25k” as your ceiling. Scope your ask to a concrete, high-leverage use.
- Use the current form (
/form/codex-for-oss/) and consider that Codex Security is gated case-by-case to high-impact/high-risk projects.