Back to all models
Get all the details on Gemini 2.5 Flash Preview, an AI model from Google. This page covers its token limits, pricing structure, key capabilities such as audio_understanding, grounding_with_search, image_understanding, available API code samples, and performance strengths.
Key Metrics
Input Limit
1M tokens
Output Limit
No data tokens
Input Cost
$0.15/1M
Output Cost
$0.60/1M
Sample API Code
import google.generativeai as genai
# Configure API key (replace with your key)
genai.configure(api_key="YOUR_API_KEY")
# Initialize the model
model = genai.GenerativeModel('gemini-2.5-flash-preview')
# Generate content
response = model.generate_content("Write a short story about a robot learning to love.")
# Print the response
print(response.text)Required Libraries
google-generative-ai
@google/generative-ai
Benchmarks
| Benchmark | Score | Source | Notes |
|---|---|---|---|
1394 | lmarena | Score for gemini-2.5-flash-preview-04-17 | |
1145 | lmarena | Score for Gemini-2.5-Flash-Preview-04-17 | |
1273 | lmarena | Score for gemini-2.5-flash-preview-04-17 | |
69.93 | LiveBench | Score for Gemini 2.5 Flash Preview (2025-05-06) | |
73.47 | LiveBench | Score for Gemini 2.5 Flash Preview (2025-05-06) | |
60.33 | LiveBench | Score for Gemini 2.5 Flash Preview (2025-05-06) | |
81.80 | LiveBench | Score for Gemini 2.5 Flash Preview (2025-05-06) | |
65.53 | LiveBench | Score for Gemini 2.5 Flash Preview (2025-05-06) | |
59.43 | LiveBench | Score for Gemini 2.5 Flash Preview (2025-05-06) | |
79.02 | LiveBench | Score for Gemini 2.5 Flash Preview (2025-05-06) | |
88 | Vellum | Score from Vellum Leaderboard | |
78.3 | Vellum | Score from Vellum Leaderboard | |
51.1 | Vellum | Score from Vellum Leaderboard |
Notes
Preview model. Our first hybrid reasoning model which supports a 1M token context window and has thinking budgets. Preview models may change before becoming stable and have more restrictive rate limits. The free tier is subject to rate limits rather than specific monthly token caps.
Capabilities
audio understanding
grounding with search
image understanding
long context
multimodal input
thinking
video understanding
vision
Supported Data Types
Input Types
text
image
video
audio
Output Types
text
Strengths & Weaknesses
Exceptional at
mathematics
Good at
hybrid reasoning
long context processing
multimodal understanding
vision tasks
instruction following
text generation
Poor at
coding
web development
Additional Information
Latest Update
May 13, 2025
Knowledge Cutoff
No data