Activefrontiermultimodal Proprietary

Gemini 1.5 Pro 002

by Google· Released September 2024· Cutoff Early 2024

Gemini 1.5 Pro 002 is Google's most capable general-purpose model, featuring a 2 million token context window and native multimodal understanding across text, images, audio, and video. It offers improved performance and lower latency compared to its predecessor, making it suitable for complex reasoning and large-scale data analysis.

Official Site API Docs

Input cost

$1.25 per 1M tokens (up to 128K tokens), $2.50 per 1M tokens (over 128K tokens)

Output cost

$5.00 per 1M tokens (up to 128K tokens), $10.00 per 1M tokens (over 128K tokens)

Context window

2M tokens

Max output

8192 tokens

Modalities

textimageaudiovideo

License

proprietary

Capabilities

Function CallingVisionCode GenerationStreamingJSON ModeAudio UnderstandingVideo UnderstandingLong Context

Best For

Complex reasoning tasks requiring very long context windows and multimodal inputs.

Strengths

Industry-leading 2M token context window
Strong multimodal understanding (text, image, audio, video)
High performance on reasoning and coding benchmarks

Limitations

Higher cost for long context usage
Not available in all regions
May still produce occasional inaccuracies in complex reasoning

Use Cases

Analyzing entire codebases or large documents

Multimodal content understanding (e.g., video analysis)

Complex data extraction and summarization

Building advanced AI assistants with long-term memory

Scientific research and data analysis

Generating detailed reports from multiple sources

Interactive tutoring and education

Improvements Over Previous Model

Reduced latency compared to Gemini 1.5 Pro 001
Improved performance on reasoning and coding benchmarks
Lower pricing for input and output tokens
Enhanced instruction following and output quality
Better handling of long context with more accurate retrieval

Back to all models