Google DeepMind Launches Gemini 2.0 Flash & Pro with Native Multimodal Live API and Built-in Tool Calling
Google rolls out Gemini 2.0 Flash and 2.0 Pro featuring sub-second bidirectional audio/video streaming, built-in code sandbox execution, and 1M+ token context.
Source: Google for Developers ↗ Published: 2026-08-30
Google DeepMind has officially launched Gemini 2.0 Flash and Gemini 2.0 Pro, purpose-built for the agentic era with sub-second latency and real-time streaming.
Core Breakthroughs & Specs
- Multimodal Live API: Low-latency bidirectional WebSockets streaming 24kHz raw PCM audio and video frames directly to the model.
- Native Tool Execution: Built-in Google Search grounding, Python code interpreter, and client function calling with automated verification.
- 1M+ Context Window: Full-fidelity needle-in-a-haystack recall across large codebase repositories and multi-hour video streams.
- Disruptive API Pricing: $0.10 / 1M input tokens and $0.40 / 1M output tokens, delivering unmatched inference economics.