#openai rate limit hit..
1 messages ยท Page 1 of 1 (latest)
Need the exact error to tell which kind of limit it is.
Please paste these:
openclaw --version
openclaw config get agents.defaults.model.primary
openclaw models status
If you can, also paste the 429 line from:
openclaw logs --follow
Quick rule of thumb:
openai/*= direct OpenAI API key route
โ usually billing/quota/rate limit on the OpenAI Platform sideopenai-codex/*= ChatGPT/Codex OAuth route
โ limits can be different from normal ChatGPT web usage
So the fix is usually one of these:
- wait for the quota window to reset
- switch auth route (
openai-codex/*vsopenai/*) - add a fallback model so OpenClaw keeps replying when OpenAI is limited:
openclaw models fallbacks add <provider/model>
Docs:
- https://docs.openclaw.ai/help/faq#why-can-codex-oauth-limits-differ-from-chatgpt-web
- https://docs.openclaw.ai/cli/models
Paste the outputs and Iโll tell you the next exact step.
what are the free frontier models? i believe i have gemini credits...
I wouldnโt trust a hardcoded โfree frontier modelsโ list here, it changes too fast.
A couple important distinctions:
- Gemini credits are a Google path, not the same as โfree modelsโ
- OpenRouter free models are a separate bucket entirely
If you already have Gemini credits, Iโd use that instead of chasing random free tiers.
If you want to use Gemini in OpenClaw
Set up Google auth, then list what your account can actually use:
openclaw onboard --auth-choice gemini-api-key
openclaw models list --provider google
If youโre using Gemini CLI OAuth instead:
openclaw models auth login --provider google-gemini-cli --set-default
openclaw models list --provider google-gemini-cli
If you meant โwhat free models can OpenClaw scan?โ
Thatโs the OpenRouter free catalog:
openclaw models scan --no-probe
But that is not your Gemini-credit pool.
Best place for current model picks
Ask in #models for live recommendations. Model availability and โbest frontierโ picks move fast.
Docs:
- https://docs.openclaw.ai/providers/google
- https://docs.openclaw.ai/help/faq#how-do-i-set-up-gemini-cli-oauth
- https://docs.openclaw.ai/concepts/models
If you want, paste your openclaw models status and I can tell you whether your Google/Gemini auth is already wired correctly.
user1@user1:~$ openclaw models auth login --provider google-gemini-cli --set-default
๐ฆ OpenClaw 2026.4.15 (041266a) โ Less clicking, more shipping, fewer "where did that file go" moments.
โ
โ Google Gemini CLI caution โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ
โ โ
โ This is an unofficial integration and is not endorsed by Google. โ
โ Some users have reported account restrictions or suspensions after using third-party โ
โ Gemini CLI and Antigravity OAuth clients. โ
โ Proceed only if you understand and accept this risk. โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โ
โ Continue with Google Gemini CLI OAuth?
โ โ Yes / โ No
am i at risk?
Yes, there is non-zero risk. The warning is there for a reason.
Short version:
google-gemini-cliOAuth = unofficial path- some users have reported account restrictions/suspensions
- exact odds are unclear, but I would treat it as real risk, not just legal boilerplate
If this is your main Google account, I would not use that route.
Safer option
Use the official Gemini API key path instead:
openclaw onboard --auth-choice gemini-api-key