Gemini 2.5 Pro Preview
Release Date: 4/5/2025
Benchmarked by 
Gemini 2.5 Pro is Google's most advanced Gemini model capable of complex reasoning over code, math, STEM problems and analyzing large datasets using multimodal inputs including audio, images, video, and PDFs with a context window exceeding 1 million tokens.
Accuracy (Average)
62.98%
Latency (Average)
352.81s
Avg. Cost (In/Out)
1.25 / 10
Context Window
1M
Max Output Tokens
66k
Input Modality
Hyperparameter settings
Default Provider :
Google
Some benchmarks may use different provider and parameters. Please refer to the benchmark page for more information.
Temperature
1
Top P
Default
Top K
Default
Max Output Tokens
65,536
Benchmarks
Accuracy
Rankings
Academic Benchmarks
Proprietary Benchmarks (contact us to get access)