Gemini 2.5 Flash
Googleai.google.dev
Fast, low-cost multimodal model with a 1M token window.
speedlow-costlong-contextvision
- Input price
- $0.30 / 1M tokensseed
- Output price
- $2.50 / 1M tokensseed
- Context window
- 1.05M tokensseed
- Max output
- 66K tokensseed
- Output speed
- 200 tokens/sseed
- Time to first token
- 0.7sseed
- Inputs
- text, image, audio, video, pdf
- Open weights
- No
- API id
gemini-2.5-flash- Released
- Jun 17, 2025seed
Benchmarks
GPQA Diamond
78%
SWE-bench Verified
50%
MMLU-Pro
80%
AIME
72%
Arena Elon/a
Benchmark scores are placeholders until sourced.