The news
Google introduced Gemini Omni 1.2 Flash on November 5. The model supports native 15-second video generation and editing. It also delivers stronger consistency and carries a 30 percent price reduction compared with the prior version.
Context
The release targets developers who need video output without moving to slower or more expensive tiers. Earlier Flash models handled shorter clips and required workarounds for anything beyond a few seconds. The new version removes that limit while lowering the per-token rate. Both announcements describe the model as the fast, cost-effective choice inside the Gemini family, aimed at teams that generate or edit video on a regular basis.
Detail
The core changes listed in the DeepMind and Google posts are identical. Native support now covers 15-second clips for both generation and editing. Frame-to-frame consistency has improved, which reduces visible jumps between frames in longer outputs. Pricing drops 30 percent from the previous Flash tier. No additional benchmark scores or latency figures appear in either post. The model stays positioned strictly as the quicker, lower-cost option rather than a replacement for heavier Gemini variants.
Developers who build tools that call the API repeatedly will see the price change directly in their usage bills. The extension to 15 seconds means a single request can now cover sequences that previously needed multiple calls or manual stitching. Consistency gains address one of the common complaints with short-clip models, where motion or object identity would drift even across brief outputs.
Why it matters
Teams that previously split video tasks across multiple short calls or paid for a heavier model now have a single, cheaper endpoint. The 30 percent reduction and the jump to 15-second native support together lower the barrier for iterative editing and longer-form generation. Production pipelines that rely on Flash for speed can now handle more complete clips without switching tiers mid-workflow.
This matters most for smaller teams and independent developers who track token costs closely. A 30 percent cut compounds quickly when video generation runs at scale, whether for product demos, marketing assets, or internal tooling. The consistency improvement reduces the amount of post-processing or regeneration needed, which indirectly saves both time and additional API spend.
At the same time, the absence of public usage data or detailed benchmarks leaves open the question of how well the new limits hold under sustained load. Teams will need to test the 15-second window themselves to see whether quality remains stable across different prompt types and motion complexity. Until those results surface, the update functions as a clear pricing and length improvement rather than a guaranteed leap in reliability.
The move also signals Google’s continued focus on making the Flash tier competitive for video without forcing users into the higher-cost Pro or Ultra lines. For developers already invested in the Gemini API, this reduces friction when video features appear in product roadmaps. For those evaluating multiple providers, the combination of longer native clips and lower price becomes one more data point in a direct comparison of per-second economics.
---
Sources:
No comments yet