AIAPI keyPricing verified 2026-10-06
Google Gemini & Veo
Gemini text models, Gemini image generation/editing (“Nano Banana”), Veo video and Gemini speech.
AProvider definition
- Provider
- Google Gemini & Veo
- What it does
- Gemini text models, Gemini image generation/editing (incl. “Nano Banana”), Veo video and Gemini speech.
- Why we need it
- Gemini text models, Gemini image generation/editing (“Nano Banana”), Veo video and Gemini speech.
- Used in
- AI Studio (text, image, video) · AI agents · Workflows
- AI capabilities
- text, reasoning, image, video, tts
- Official site
- https://aistudio.google.com/app/apikey
BAccount requirements
- Account type
- Google account with Google AI Studio (optionally linked to a Google Cloud billing account)
- Prerequisites
- Email sign-up
- Billing set up on the provider account for production volume
CConnection method
- Authentication
- API key
- Where to enter it
- Your workspace: App → Integrations → Providers & API keysWorkspace owner or admin. Keys added here are used only by your workspace (AI keys here don't use plan credits).
- Platform-wide (Super Admin): Super Admin → Platform → IntegrationsPlatform owner. Used for every customer who hasn't added their own key; can also be set as PLATFORM_KEY_<PROVIDER> on the API server.
Setup steps
- 1Sign in to the provider console and open aistudio.google.com → Get API key.
- 2Create a new API key. Give it a name such as “SocialMati production”.
- 3Copy the key immediately — most providers show it only once.
- 4Add billing or credits on the provider account if your plan needs it.
- 5Paste the key in the application (see “Where to enter it”), save, then click Test.
DPermissions & API scopes
No OAuth scopes — access is controlled by the key or token itself. Restrict keys to the APIs listed in the setup steps where the provider allows it.
This list is generated from the scopes the application's code requests.
EFree vs paid
- Free account / free API access
- Yes — a free tier with low per-model rate limits (requests per minute/day); Pro models are paid-only for new users since April 2026.
- Card or billing account required?
- Not for the free tier. Paid tier: link a billing account and prepay at least $5.
- Paid plan / costs
- Usage-based per token / image / second of video; spending caps apply.
- Quota and usage limits
- Per-model RPM, TPM and RPD; free-tier prompts may be used to improve Google products.
Checked against official/current sources on 2026-10-06. Provider pricing changes often — re-check before publishing.
FSetup screenshots
Capture each step with real values masked. Save images as public/docs/integrations/gemini/1.png … 10.png and they appear here automatically.
- Screenshot to add1. Account login / creationThe provider's sign-in or sign-up page.
- Screenshot to add2. Developer consoleThe developer console or dashboard home.
- Not applicable3. Project / app creationCreating the app or project used for this integration.
- Not applicable4. API enablementThe API or product enabled for the project.
- Not applicable5. OAuth configurationConsent screen / OAuth settings with the redirect URIs.
- Screenshot to add6. API key / credentials creationCreating the key, secret or token (mask the value).
- Not applicable7. Permissions / scopesThe scopes or permissions requested or approved.
- Screenshot to add8. Billing / quota settingsBilling, credits or quota page.
- Screenshot to add9. Application connectionThe provider card in our application with fields filled (secrets masked).
- Screenshot to add10. Successful connectionThe green Test result or the connected account listed.
GApplication configuration
Never include real API keys, client secrets, access tokens or passwords in screenshots, tickets or documentation. Values below are field names and examples only.
| Field | Required | Example / default | Notes |
|---|---|---|---|
| API key | Yes | •••••••• | Stored encrypted; shown masked after saving. |
HTesting
- 1Click Test on the Gemini card — it shows how many models the key can access.
- 2Generate a short piece of content in AI Studio with this provider's model selected (Model selector → choose a model from this provider).
ITroubleshooting
| Problem | What to do |
|---|---|
| 401 / invalid API key | The key was revoked, mistyped or belongs to another project. Create a new key and paste it again. |
| 402 / insufficient credits / quota exceeded | Add credits or raise the spending limit on the provider account. |
| “No text model is available” | The provider is switched off or limited to certain plans in Super Admin → AI Management, or its models are disabled. The error message names the cause. |
| Invalid credentials | The key or token is wrong, revoked or from another project. Create a new one and paste it again, then click Test. |
| Missing permissions | The key's project or role doesn't include this API. Grant access or create the key in the right project. |
| API not enabled | Enable the API/product for the project or app in the provider console and wait a few minutes. |
| Quota exceeded | Wait for the quota window to reset (see section E) or request a higher quota / add credits. |
| Account not eligible | Check the account requirements in section B (e.g. business/professional account, verified property, registered organisation). |
| Billing required | Add a payment method or prepaid credits on the provider account (see section E). |
JDisconnect & security
- Revoke access at the provider
- Delete the key in the provider console, then remove it from the provider card in the application.
- Disconnect in the application
- App → Integrations → Providers & API keys or Super Admin → Platform → Integrations → open the card → Remove / Disconnect. Connected accounts can be disconnected per brand.
- Rotate credentials
- Create the new key or secret first, paste it into the same card and Test, then delete the old one at the provider — no downtime.
- How credentials are stored
- Encrypted at rest with AES-256-GCM using the server's ENCRYPTION_KEY, shown masked after saving, never returned to the browser, and used only from the API server.
- What data the application can access
- Only the prompts and files you send for generation. Requests are sent from our server; the key is never exposed to the browser.

