MODELS
Gemini 2.5 Flash
Cheap, fast multimodal workhorse with a 1M-token window.
Specs
- Context window
- 1,000,000
- Max output
- 65,536
- Modalities
- text, image, audio, video
- Tool use
- ✓
- Vision
- ✓
- Streaming
- ✓
- License
- proprietary
- Released
- 2025-06-01
Pricing
- Input / 1M
- $0.30
- Output / 1M
- $2.50
- Cached input / 1M
- $0.07
Cost estimate
Google's mid-tier Gemini model built for high-volume, latency-sensitive workloads. Handles text, images, audio, and video natively, with tool use and streaming, across a 1M-token context. At $0.3/M input and $2.5/M output it sits well below GPT-5 mini and Claude Haiku on price while keeping multimodal parity. Aimed at builders running RAG, classification, transcription, or video understanding at scale.
Editor's verdict
Pick this when you need to chew through long documents, audio, or video on a budget — its 1M context and native video input are still rare at this price point. Reasoning depth lags behind Gemini 2.5 Pro, GPT-5, and Claude Sonnet 4.5, so don't reach for it on hard math, complex agents, or nuanced code review. The sweet spot is high-volume multimodal pipelines where "good enough and fast" beats "best".
Reviews
No reviews yet. Be the first.
Last updated: 2026-04-29