Gemini family — multimodal models with massive context windows and hybrid reasoning.
Available models
12
Starting rate
60 RODI / 1M
Max context
2M
Most intelligent model built for speed. Combines frontier intelligence with superior search and grounding. Latest model launched at Google I/O May 2026. Thinking tokens included in output price. via Google Gemini API (direct).
Price
Most capable model in Gemini lineup. 2M token context window — largest in industry. Tiered pricing: input/output roughly doubles above 200K tokens. Paid tier only, no free access. Preview status: may change before GA, more restrictive rate limits. Supports multimodal understanding, agentic capabilities, and coding. via Google Gemini API (direct).
Price
Most cost-efficient model in Gemini API. Moved from preview to GA on May 7, 2026. Optimized for high-volume agentic tasks, translation, and simple data processing. Audio input priced 2x higher than text/image/video. Flat pricing regardless of context length. via Google Gemini API (direct).
Preview version of Flash-Lite. Same pricing as GA version but with more restrictive rate limits and no stability guarantees. via Google Gemini API (direct).
Previous flagship model. Tiered pricing above 200K tokens. Batch pricing 50% off. Context caching read at $0.31/MTok. via Google Gemini API (direct).
Price
Most affordable model in Gemini 2.5 family. Recommended entry point for high-volume tasks. Audio input 3x more expensive than text. Being superseded by Gemini 3.1 Flash-Lite. via Google Gemini API (direct).
Popular mid-tier model from previous generation. Audio input 3.33x more expensive than text. Being superseded by Gemini 3.x models. via Google Gemini API (direct).
Imagen 4 Ultra — highest fidelity. $0.06 per generated image (Gemini Developer API). via Google Gemini API (direct).
Price
Imagen 4 Standard — default Imagen 4 tier. $0.04 per generated image (Gemini Developer API). via Google Gemini API (direct).
Price
Latest video generation model. Available to developers on paid tier only. Preview status: may change before GA. Accessible via the Gemini API. via Google Gemini API (direct).
Price
Imagen 4 Fast — lower latency image generation. $0.02 per generated image (Gemini Developer API). via Google Gemini API (direct).
Price
Standard embedding model for vector generation. Free tier available. Used for semantic search, RAG, and similarity tasks. Grounding with Google Search available on top: $0.15/MTok embedding tokens plus standard retrieval costs. via Google Gemini API (direct).
Price
