| | # Replace YOUR_PROJECT_ID with your Google Cloud project ID and | | | # update each --preferred-value from its default to your target daily limit. | | | # 1. Daily input tokens per project (Default: 200B) | | | gcloud quotas preferences update custom-GenAiInputTokensPerDay \ |
| | --project=YOUR_PROJECT_ID \ |
| | --service=bigquery.googleapis.com \ |
| | --quota-id=GenAiInputTokensPerDay \ |
| | --preferred-value=200000000000 \ |
| | --allow-high-percentage-quota-decrease \ |
| | --allow-missing |
| | # 2. Daily input tokens per user (Default: 150B) |
| | gcloud quotas preferences update custom-GenAiInputTokensPerUserPerDay \ |
| | --project=YOUR_PROJECT_ID \ |
| | --service=bigquery.googleapis.com \ |
| | --quota-id=GenAiInputTokensPerUserPerDay \ |
| | --preferred-value=150000000000 \ |
| | --allow-high-percentage-quota-decrease \ |
| | --allow-missing |
| | # 3. Daily output tokens per project (Default: 20B) |
| | gcloud quotas preferences update custom-GenAiOutputTokensPerDay \ |
| | --project=YOUR_PROJECT_ID \ |
| | --service=bigquery.googleapis.com \ |
| | --quota-id=GenAiOutputTokensPerDay \ |
| | --preferred-value=20000000000 \ |
| | --allow-high-percentage-quota-decrease \ |
| | --allow-missing |
| | # 4. Daily output tokens per user (Default: 15B) |
| | gcloud quotas preferences update custom-GenAiOutputTokensPerUserPerDay \ |
| | --project=YOUR_PROJECT_ID \ |
| | --service=bigquery.googleapis.com \ |
| | --quota-id=GenAiOutputTokensPerUserPerDay \ |
| | --preferred-value=15000000000 \ |
| | --allow-high-percentage-quota-decrease \ |
| | --allow-missing |
source & further reading
gist.github.com — original article
Caffeinate AI: Keep macOS awake while one or more monitored apps are running (for long agentic sessions with ChatGPT/Codex or Claude)
AGENTS.md
Subagent guidance for Codex (copy the raw contents into your ~/.codex/AGENTS.md)