Gemini 2.5 Pro is Google's most advanced AI model, engineered for deep reasoning and thoughtful response generation. It outperforms on key benchmarks, demonstrating exceptional logic and coding proficiency. Optimized for building dynamic web applications, autonomous code systems, and code adaptation, it delivers high-level performance. With built-in multimodal capabilities and an extended context window, the model efficiently processes large datasets and integrates diverse information sources to tackle complex challenges.
| Gemini 2.5 Pro | Claude Sonnet 4 | |
|---|---|---|
Web Site
| ||
Provider
| ||
Chat
| ||
Release Date
| ||
Modalities
| text images voice video | text images |
API Providers
| Google AI Studio, Vertex AI, Gemini app | Anthropic API, Amazon Bedrock, Google Cloud's Vertex AI |
Knowledge Cut-off Date
| - | Unknown |
Open Source
| No | No |
Pricing Input
| Not available | $3 per million tokens |
Pricing Output
| Not available | $15 per million tokens |
MMLU
| Not available | 86.5% Source |
MMLU-Pro
| Not available | - |
MMMU
| 81.7% Source | 74.4% Source |
HellaSwag
| Not available | - |
HumanEval
| Not available | - |
MATH
| Not available | - |
GPQA
| 84.0% Diamond Science Source | 75.4% Diamond Source |
IFEval
| Not available | - |
SimpleQA
| 52.9% | - |
AIME 2024 | 92.0% | - |
AIME 2025 | 86.7% | 75.5% Source |
Aider Polyglot
| 74.0% / 68.6% | - |
LiveCodeBench v5
| 70.4% | - |
Global MMLU (Lite)
| 89.8% | - |
MathVista
| - | - |
Mobile Application | ||
VideoGameBench | ||
Total score | 0.48% | - |
Doom II | 0% | - |
Dream DX | 4.8% | - |
Awakening DX | 0% | - |
Civilization I | 0% | - |
Pokemon Crystal | 0% | - |
The Need for Speed | 0% | - |
The Incredible Machine | 0% | - |
Secret Game 1 | 0% | - |
Secret Game 2 | 0% | - |
Secret Game 3 | 0% | - |
MathArena | ||
| Avg. Score | 81% | - |
AIME 2025 A test based on problems from the American Invitational Mathematics Examination, designed to assess the mathematical skills of models. | 87% | - |
HMMT February 2025 A test based on problems from the Harvard-MIT Mathematics Tournament, February 2025, designed to assess the mathematical skills of models. | 82% | - |
BRUMO 2025 | 90% | - |
SMT 2025 A test based on problems from the Stanford Math Tournament, 2025, designed to assess the mathematical skills of models. | 85% | - |
CMIMC 2025 A test based on problems from the Canadian Mathematical Olympiad, 2025, designed to assess the mathematical skills of models. | 58% | - |
Compare AI. Test. Benchmarks. Mobile Apps Chatbots, Sketch
Copyright © 2026 All Right Reserved.
Comments (1)
Mazen
11 August 2025Good program