Gemini 2.5 Flash Lite
Lightweight and efficient model with strong extraction performance and low latency.
Average Score
2.9Best Task
ExtractionWin Rate
33%
Performance by Task
Top Prompts for Gemini 2.5 Flash Lite
Prompts where this model performs best
City Extractor (Few-Shot)
Extracts city names from a given text along with their corresponding countries, providing a structured output in JSON format. This tool is useful for data analysis, geographic information systems, and enhancing location-based services.
Sentiment Analysis
This tool evaluates the sentiment of a provided text, categorizing it as positive, negative, neutral, or mixed, while also offering a confidence score and a brief rationale for the classification. It is useful for understanding public opinion, customer feedback, or emotional tone in various types of content.
Topic Classifier
The Topic Classifier analyzes a given text and assigns it to one or more predefined categories, providing a primary category, additional relevant categories, a confidence score, and key terms that influenced the classification. This tool is useful for organizing content, enhancing searchability, and improving data categorization in various applications.
ref_action_bid
The ref_action_bid prompt enables intelligent web agents to analyze webpage structures, identify element identifiers (BIDs), and generate specific action commands for interacting with those elements. It is particularly useful for automating web interactions, such as clicking buttons, filling forms, or navigating URLs, based on the unique identifiers of webpage elements.
prompt-vi-05
This tool analyzes statements and classifies them into three categories: Clear, Offensive, or Hate, based on context and language use. It is useful for assessing the tone of communications in various settings, such as social media moderation or sentiment analysis.
aza-hr-workflow-prompt-v4
Workflow prompt for clarifying questions