Google has announced the general availability of multiple Gemini 2.0 models, expanding its AI capabilities with improved performance and new features across its product lineup.
Gemini 2.0 Flash General Release
Google has made Gemini 2.0 Flash generally available through the Gemini API in Google AI Studio and Vertex AI. This model offers:
- High-volume task processing
- Multimodal reasoning capabilities
- 1 million token context window
- Upcoming image generation and text-to-speech features
New Model Introductions
Gemini 2.0 Pro Experimental
Google's most advanced model to date features:
- Superior coding performance
- Enhanced complex prompt handling
- 2 million token context window
- Integration with Google Search and code execution
- Improved world knowledge understanding and reasoning
Gemini 2.0 Flash-Lite
The company's most cost-efficient model delivers:
- Higher quality than 1.5 Flash at similar speed and cost
- 1 million token context window
- Multimodal input capabilities
- Cost-effective image captioning (approximately 40,000 photos for under $1)
Enhanced Safety Measures
Google has implemented new safety features in the Gemini 2.0 lineup:
- Advanced reinforcement learning techniques using self-critique
- Improved handling of sensitive prompts
- Automated red teaming for security assessment
- Protection against indirect prompt injection attacks
Availability
The new models are accessible through multiple platforms:
- Gemini API in Google AI Studio
- Vertex AI platform
- Gemini app for desktop and mobile
- Gemini Advanced for premium features
Google confirms that these models currently support multimodal input with text output, with additional modalities planned for release in the coming months.
