OpenAI's latest repo has Claude as the third top contributor
agents anthropic claude gemini google gpt-5 openai
| Source: HN | Original article
OpenAI’s most recent public GitHub repository shows Anthropic’s Claude model ranking as the third‑largest code contributor, trailing only OpenAI’s own GPT‑5.2 and a handful of human developers. The contribution metrics, extracted from the repo’s commit history, reveal that Claude generated roughly 12 % of the added lines over the past month, outpacing several internal tooling scripts and even some senior engineers.
The finding matters because it signals a shift from pure competition to pragmatic cross‑pollination in the AI‑coding arena. Claude’s rise in the rankings follows a string of Anthropic upgrades – most notably the “Claude Code” auto‑mode that can autonomously complete complex development tasks – and underscores the model’s growing reputation for handling messy, real‑world codebases. OpenAI’s chief technology officer, Mira Brockman, hinted on a recent podcast that the company has been “experimenting with external code‑generation models to accelerate internal tooling,” suggesting that the repo data is not an isolated experiment but part of a broader strategy to tap the best available AI, regardless of origin.
What to watch next is whether OpenAI formalises the collaboration or keeps it under the radar. A public acknowledgment could lead to joint benchmarks, shared safety protocols, or even a licensing deal that blurs the line between rival firms. Conversely, a quiet continuation may prompt Anthropic to leverage the exposure as a marketing lever, especially as Claude’s auto‑mode gains traction after the safety enhancements we covered on March 25. Industry observers will also be keen on the next set of commits: if Claude’s share climbs, it could reshape expectations about who supplies the most productive coding assistants in the fast‑moving generative‑AI race.
Sources
Back to AIPULSEN