API gratuiteLLMClé APIOfficiel
Google Gemini API
Free multimodal Gemini Flash with a daily request budget, no card
Google's Gemini API offers a free tier through Google AI Studio covering the Gemini Flash and Flash-Lite model families, including text, vision, and long-context use. You generate a key in AI Studio with no credit card and call a REST endpoint or use the official SDKs.
Palier gratuit
Free tier on Flash / Flash-Lite models (no card). Pro models moved behind a paid (billing-enabled) project in 2026. Free-tier limits are not guaranteed; Google cut them 50-80% on 7 Dec 2025 and revises them without notice.Limites de débit
Per-project, per-model (verified June 2026): Gemini 2.5/3 Flash = 10 RPM, 250 req/day; Flash-Lite = 15 RPM, 1,000 req/day; shared 250K TPM. Pro (5 RPM, 100 RPD) now needs billing. Daily quotas reset midnight Pacific. Check AI Studio for your project's live limits.Authentification
Clé APIURL de base
https://generativelanguage.googleapis.com/v1betaCe que vous pourriez construire avec
- Multimodal automations that read images, PDFs, or audio alongside text
- Long-context document summarization on Flash's large token window
- Free embeddings for lightweight RAG prototypes
- Vision-enabled Claude Skill fallbacks where you want Google's multimodal coverage
Endpoints clés
- POST
/models/{model}:generateContentGenerate text/multimodal content from a prompt - POST
/models/{model}:streamGenerateContentStream a multimodal generation - POST
/models/{model}:embedContentProduce text embeddings - GET
/modelsList available models and supported methods
Members
Members get the runnable recipe
Sign in free to copy the example request, see a sample response, and read the gotchas.
Tags
llmmultimodalvisionembeddingslong-context