Gemini 2.5 Flash
Multimodal
by Google DeepMind
+
+
+
+
About
Gemini 2.5 Flash, released by Google in June 2025, is a large language model from the Gemini 2.5 family optimized for high-throughput, cost-efficient deployments with multimodal reasoning. It features a 1M token context window, hybrid thinking control, and native support for text, image, video, and audio input. Gemini 2.5 Flash targets latency-sensitive applications, document analysis, and high-volume API workloads that benefit from combined reasoning and generation in a single model.
+
+
+
+
Pricing Range
Input (per 1M)$0.15 -$0.15
Output (per 1M)$0.60 -$0.60
Providers2
+
+
+
+
Timeline
ReleasedJun 17, 2025
+
+
+
+
Specifications
Capabilities
Multimodal
+
+
+
+
License & Family
License
Proprietary
Performance Overview
Performance metrics and category breakdown
Overall Performance
6 benchmarks
Average Score
18.6%
Best Score
78.3%
High Performers (80%+)
0Performance Metrics
Max Context Window
1.0MTop Categories
Science
78.3%
Coding
16.4%
Reasoning
12.1%
Tool Use
3.4%
Agents
2.5%
+
+
+
+
All Benchmark Results for Gemini 2.5 Flash
Complete list of benchmark scores with detailed information
| GPQA Diamond | Science | 78.30 | 78.3% | Unverified | |
| Terminal Bench 2.0 | Coding | 16.40 | 16.4% | Unverified | |
| Humanity's Last Exam | Reasoning | 12.08 | 12.1% | Unverified | |
| MCP-Atlas | Tool Use | 3.40 | 3.4% | Unverified | |
| BrowseComp | Agents | 2.50 | 2.5% | Unverified | |
| MMMU | Multimodal | -1.00 | -1.0% | Unverified |