OpenClaw rebuilt model discovery so new models ship without waiting for a release
heyneighbor · x · 2026-11-07
An engineering deep-dive on how OpenClaw eliminated release-gated model support. Previously, each provider plugin shipped a model manifest inside the package, so when OpenAI launched GPT-6.1 Sol, ChatGPT sign-in users couldn't select it until OpenClaw released a new version bundling a newer Codex.
Over 5 weeks and 80 PRs, the team rebuilt model discovery around three sources:
- Self-updating catalog: catalog.openclaw.ai rebuilds every 4 hours from models.dev data plus their own manifests and pricing, holding 1,000 models; a running Gateway hot-switches to newer catalogs mid-session while in-progress turns keep their starting catalog for consistent models/prices.
- Account-based discovery: reading available models from the user's provider account.
- Codex versioning: detecting what's installed locally.
The post also covers a fix they shipped and reverted, and what still requires a release.
More from coding & agent
- Developer rebuilds hospital notes as one document with a version-history timeline — BraydonDymm · 2026-11-07
- Aside launches Computer Use agent: 2.7x faster than Codex with 65% fewer tokens — ycombinator · 2026-11-07
- Regression testing with AI coding: fixing one thing quietly breaks another — _ACTUALin · 2026-11-07
- Together AI's TKCC lets agents hot-patch inference engines: 252x more experiments, 2.6x throughput — simran_s_arora · 2026-11-07
- Every's Slack cloned with 51 AI coworkers built from team messages — every · 2026-11-07
- New MCP Agent Usability leaderboard ranks how well coding agents use MCP servers — alejxb · 2026-11-07