Gemini Omni 1.2 Flash

Google introduced Gemini Omni 1.2 Flash, a Flash-tier multimodal model positioned as the low-latency, low-cost option for high-volume inference, according to the company's developer blog. The model makes native, bidirectional audio and video streaming a core API capability, which Google says bypasses separate speech-to-text and text-to-speech pipelines and enables sub-second voice-to-voice agent interactions through a single unified API call. Google advises teams running Flash models in production to re-benchmark latency, token pricing and multimodal handling before swapping the new model into existing pipelines.

Hacker News https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemini-omni-1-2-flash/ Gemini Omni 1.2 Flash Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated. A new Gemini Flash-tier multimodal "Omni" model is rolling out, positioned as the low-latency, low-cost option in the lineup for high-volume inference. If you're running Flash models in production, expect to re-benchmark latency, token pricing, and multimodal handling before swapping it into existing pipelines, since Flash-tier upgrades typically shift cost/quality tradeoffs rather than just adding capability. Gemini Omni 1.2 Flash establishes native, bidirectional audio and video streaming as a core API capability, bypassing the latency and cost of separate speech-to-text and text-to-speech pipelines. For production LLM applications, this enables sub-second voice-to-voice agent interactions through a single, unified API call. This eliminates complex audio orchestration middleware and shifts your engineering challenge entirely to managing real-time session state and token consumption during live interactions.