Gemini API

Hosts: generativelanguage.googleapis.com · 17 facts

Error meanings

• ✅ verified:server:prod · model is not found: models/gemini-agent-facts-nosuch for api version v#beta · GET /v1beta/models/{model} · env prod
• ✅ verified:server:prod · * listmodelsrequest.page_size: page_size must be positive · GET /v1beta/models · env prod
• ✅ verified:server:prod · invalid json payload received. unknown name "agent_facts_unknown": cannot bind query parameter. field 'agent_facts_unknown' could not be found in request message. · GET /v1beta/models · env prod

Rate limits

• 🔵 official · Requests-per-day quotas reset at midnight Pacific time, not at midnight UTC. · POST /v1beta/models/{model_and_method} · env prod
• 🔵 official · Rate limits (requests per minute, input tokens per minute, requests per day) apply per Google Cloud project, not per API key: several keys of one project share them. · POST /v1beta/models/{model_and_method} · env prod
• 🔵 official · Exceeding a rate limit or quota returns HTTP 429 RESOURCE_EXHAUSTED, a transient error, like 503 UNAVAILABLE; 400, 402 and 403 are client errors that a retry does not fix. · POST /v1beta/models/{model_and_method} · env prod

Silent behavior

• ✅ verified:server:prod · Errors carry the google.rpc envelope: error.code is the HTTP status as a number, error.status a name such as INVALID_ARGUMENT, and error.details[] holds objects typed by @type (e.g. google.rpc.BadRequest). · GET /v1beta/models · env prod
• ✅ verified:server:prod · model is not found: models/gemini-flash-latest for api version v# · GET /v1/models/{model} · env prod

Auth quirks

• ✅ verified:server:prod · api key not valid. please pass a valid api key. · GET /v1beta/models · env prod
• 🔵 official · your api key was reported as leaked. please use another api key. · POST /v1beta/models/{model_and_method} · env prod
• ✅ verified:server:prod · method doesn't allow unregistered callers (callers without established identity). please use api key or other form of api consumer identity to call this api. · GET /v1beta/models · env prod

Pagination

• ✅ verified:server:prod · invalid page_token. · GET /v1beta/models · env prod
• ✅ verified:server:prod · GET /v1beta/models without pageSize returns 50 models and a nextPageToken; the rest of the list is only reachable with pageToken. · GET /v1beta/models · env prod
• ✅ verified:server:prod · pageSize=0 on GET /v1beta/models is not refused: it returns the default page of 50 models with a nextPageToken. · GET /v1beta/models · env prod

Deprecations

• ✅ verified:server:prod · model is not found: models/gemini-#.#-flash for api version v#beta · GET /v1beta/models/{model} · env prod
• ✅ verified:server:prod · gemini-2.0-flash, whose earliest shutdown date on the deprecations page is June 1, 2026, still answers GET /v1beta/models/gemini-2.0-flash with HTTP 200 and version 2.0. · GET /v1beta/models/{model} · env prod

Schema drift

• ✅ verified:server:prod · A model object from GET /v1beta/models/{model} has 12 top-level keys (description, displayName, inputTokenLimit, maxTemperature, name, outputTokenLimit, supportedGenerationMethods, temperature, thinking, topK, topP, version). · GET /v1beta/models/{model} · env prod

Data: CC BY 4.0 · apifacts registry · This page is generated; edit the YAML, not the page.