Can I get Google Gemini spend per day from the API?

No. Count calls and price them yourself, or let the Argus SDK do it.

What happens when Google Gemini credits run out?

429 RESOURCE_EXHAUSTED on any limit/guardrail or exhausted credits. Auto-recharge is available.

LLMs

How to get your Google Gemini API usage, balance and spend programmatically

The Gemini API has no balance or usage endpoint. Spend lives in Google Cloud Billing, not the Gemini API, and responses carry no rate-limit headers. Per-call token counts are in each response's usageMetadata.

What the Google Gemini API exposes

Signal	Google Gemini
Balance	No
Spend history	No
Cost breakdown	No
Plan and quota	No
Cost on each response	Token counts only
Rate-limit headers	None
Credential	API key

Call it yourself

Credential. A Gemini API key. Argus only validates it, with GET /v1beta/models.

What each response carries

{
  "responseId": "c2b1…",
  "modelVersion": "gemini-2.5-flash",
  "usageMetadata": {
    "promptTokenCount": 1204,
    "candidatesTokenCount": 311,
    "cachedContentTokenCount": 0,
    "thoughtsTokenCount": 128
  }
}

Trimmed to the fields Argus reads. Values are examples.

No billing API, so measure from your side

Argus measures Gemini from your side: the SDK records usageMetadata token counts on every call, and Argus prices them against its per-model rates. For billed totals, use the Cloud Billing export to BigQuery.

Gotchas

429 RESOURCE_EXHAUSTED covers rate limits, spend guardrails and exhausted credits alike.
Prompts over 200K tokens reprice higher on long-context models.
Thinking tokens (thoughtsTokenCount) bill as output. Count them or your estimate runs low.

How Google Gemini charges

Free tier plus paid per-token (prepaid credits, or invoiced at Tier 3). Long-context prompts (>200K) reprice higher. No usage/billing API and no rate-limit headers.

Tier 1: ~$250/mo cap; $10 per 10-min guardrail
Tier 3: $20k–100k+ cap; postpay unlocked

Free allowance. 1 a day. Flash-class models at low RPM/RPD; Pro-class removed from free tier.

At zero. 429 RESOURCE_EXHAUSTED on any limit/guardrail or exhausted credits.

requests per minute, per model. Per-model × per-tier RPM/TPM/RPD; live dashboard is source of truth. No rate-limit headers.

Or let Argus track it

Paste the key once. Argus keeps the ledger, and your coding agent reads it through the MCP server.

With the SDK, records token counts on each response and prices them per model, so you see what a call cost before the provider's daily bucket lands.
Emails you on low balance, days to zero, a spend spike, or projected overage.
Hands all of it to your coding agent over MCP, so it can ask what a run cost or how much runway is left without leaving the editor.

> what has google_gemini cost today, and how long until it runs out?

get_overview       → google_gemini: mtd spend, burn per day, days to zero
get_request_costs  { "provider": "google_gemini", "limit": 25 }
get_advisories     → credits_low, spend_spike, rate_limit_low …

Questions

Does Google Gemini have an API for its balance?: The Gemini API has no balance or usage endpoint. Spend lives in Google Cloud Billing, not the Gemini API, and responses carry no rate-limit headers. Per-call token counts are in each response's usageMetadata.
Can I get Google Gemini spend per day from the API?: No. Count calls and price them yourself, or let the Argus SDK do it.
What happens when Google Gemini credits run out?: 429 RESOURCE_EXHAUSTED on any limit/guardrail or exhausted credits. Auto-recharge is available.
What key does Argus need for Google Gemini?: A Gemini API key. Argus only validates it, with GET /v1beta/models.

Every balance endpoint

The balance call for each provider, side by side.