Gemini 2.5 Pro
FrontierGoogle•Released on 2025-03-25
Google's state-of-the-art reasoning model with "thinking" capabilities (experimental preview March 2025, GA June 2025). 1M context, native multimodal (text, image, audio, video). Excels at math, science, coding, and complex problem-solving. Great value at $1.25/$10.
87
Overall Score
Voice of the community
sample 1,456“Gemini 2.5 Pro is state-of-the-art across benchmarks requiring advanced reasoning. Leads in math and science benchmarks like GPQA and AIME 2025.”
“The multimodal capabilities are unmatched. Being able to input video directly is a game changer for my workflow.”
“For pure coding tasks I still prefer Claude, but Gemini 2.5 Pro is my go-to for research and analysis.”
Core Specs
1049K
Context Window
66K
Max Output
ReasoningOpen Sourcetextimageaudiovideofile
Scenario Scores
Pros & Cons
Sentiment75% +18% ·7% −
Pros
- +Top-tier LMArena performance
- +True multimodal (text, image, audio, video)
- +1M context window
- +Built-in thinking/reasoning mode
- +Excellent math and science performance
Cons
- −Reasoning mode increases latency
- −Not as strong as Claude for coding
- −Google ecosystem dependency
Pricing
Input (per 1M tokens)$1.25
Output (per 1M tokens)$10.00
Subscription$20/month
Free trial available
Updated on 2026-03-07
Get Started with Gemini
Benchmarks
gpqa84.5%
aime202578%