Comprehensive side-by-side LLM comparison
Gemini 3 Pro leads with 29.5% higher average benchmark score. Gemini 3 Pro offers 3.1K more tokens in context window than Nova 2 Lite. Nova 2 Lite is $22.10 cheaper per million tokens. Overall, Gemini 3 Pro is the stronger choice for coding tasks.
Google DeepMind
Gemini 3 Pro, released by Google in November 2025, is a large language model from the Gemini 3 family designed for complex reasoning, broad world knowledge, and multi-step problem solving. It features a 1M token context window, 64K maximum output tokens, and native multimodal input including text, images, audio, and video. Gemini 3 Pro targets research-grade reasoning, advanced data analysis, and applications requiring comprehensive understanding across diverse input formats.
Amazon
Amazon Nova 2 Lite, released by Amazon Web Services on December 2, 2025, is a fast, cost-efficient reasoning model available on Amazon Bedrock with a 1M token context window enabling extended document, video, and image analysis. It features three extended thinking intensity levels, built-in code interpreter, web grounding tools, and native support for text, image, video, and document input. Nova 2 Lite targets cost-sensitive agentic applications, document analysis pipelines, and real-time workloads requiring multimodal reasoning.
14 days newer

Gemini 3 Pro
Google DeepMind
2025-11-18

Nova 2 Lite
Amazon
2025-12-02
Cost per million tokens (USD)
Gemini 3 Pro
Nova 2 Lite
Context window and performance specifications
Average performance across 1 common benchmarks
Gemini 3 Pro
Nova 2 Lite
Performance comparison across key benchmark categories
Gemini 3 Pro
Nova 2 Lite
Available providers and their performance metrics
Gemini 3 Pro
Google Cloud Vertex AI
Nova 2 Lite
Gemini 3 Pro
Nova 2 Lite
Gemini 3 Pro
Nova 2 Lite
AWS Bedrock