Gemini Omni 1.2 Flash: 30% price cut, 15-second video

What changed

Google released Gemini Omni 1.2 Flash on November 5, 2026, with a 30% price reduction and support for native 15-second video clips. The pricing model now sits at $0.07 per second for 720p video, compared to roughly $0.10 per second for Omni 1.1 Flash. A full 15-second clip at 720p costs about $1.05 to generate.

The model is available on the Gemini API and Google Cloud. Google is rolling out access to Google Vids, Google Ads, and the Gemini app for business users.

Six core upgrades ship with this version:

• Native clip length extended to 15 seconds with duration control in one-second increments from 3 to 15 seconds
• Reference adherence improved when combining character, product, or style images with motion clips up to 3 seconds to maintain consistency across camera angles
• Stronger consistency for faces, garments, packaging text, and brand colors across transitions with reduced morphing
• More natural dialogue featuring clearer lip-sync and distinct multi-speaker conditioning
• More controllable camera movements for flythroughs, zoom-outs, and orbits with fewer jump cuts
• Multi-turn conversational editing that spans full 15-second clips without regeneration

Google emphasizes a three-step cost efficiency pattern: draft at lower resolution (360p or 720p), edit specific details via follow-up prompts instead of re-rendering the entire clip, then upscale the final result to 1080p or 4K. This workflow approach reduces iteration costs compared to generating each variation from scratch.

Why it matters

The pricing reduction reshapes unit economics for teams evaluating AI video tools. At $0.07 per second, a 15-second commercial spot now costs approximately $1.05. Doubling the native clip length from 10 to 15 seconds eliminates the need to stitch multiple segments or use external tools to extend footage.

Multi-turn editing spanning full clips without regeneration addresses a key production bottleneck. Previously, changing a background or swapping a product required re-rendering the entire clip. Now a single conversation thread enables rapid production of ad variants for different platforms, audiences, or campaigns from one base render. This approach reduces both cost and iteration time.

For video agencies, marketing teams, and e-commerce platforms, the combination of lower per-second costs and longer native clips increases output per budget dollar. Improvements to consistency also matter practically. If faces, logos, and product details hold across camera movements, fewer renders get rejected during review, minimizing wasted generation credits.

The source notes that Google has not published benchmark scores for version 1.2. Vendor claims rest on side-by-side demonstrations and feedback from early partners. Google acknowledges that fast sports action, rapid script changes, and complex physics remain areas needing improvement.

What to test

Teams considering adoption should verify these vendor claims with their own content:

• Consistency across longer clips. Test with your brand logos, product shots, and character assets. Does the model maintain text, colors, and proportions across full 15-second clips, or does visual drift still occur on extended renders?

• Multi-turn editing workflow. Confirm that editing via the previous_interaction_id parameter does not re-charge for the base video or introduce latency in the editing process.

• Upscaling quality versus native generation. Draft at 360p or 720p, then upscale to 1080p or 4K. Compare the resulting visual fidelity and file size against generating directly at target resolution.

• Speech and dialogue quality. Generate multiple 15-second clips featuring multi-character dialogue and evaluate lip-sync accuracy, speaker separation, and overall naturalness relative to Omni 1.1 samples if available.

• Camera movement stability. Test complex motions such as continuous drone flights or orbits. Look for jump cuts, scale shifts, or unwanted morphing that Google reports as reduced but not eliminated.

• Cost per variant in production workflow. Run a batch of ad variants using multi-turn editing and track total spend versus regenerating each variant independently from scratch.

• API rate limits versus provisioned throughput. Determine whether standard API rate limits accommodate your pipeline or if you need reserved capacity and associated pricing.

The conclusion

Gemini Omni 1.2 Flash represents a meaningful update addressing friction points for teams producing ad content, product videos, and branded material. The 30% pricing reduction combined with 15-second native clips and multi-turn editing tackle real production constraints. The consistency and speech improvements matter if they hold in real workflows.

The primary risk is the absence of published benchmarks for version 1.2. Early claims derive from demonstrations and self-selected partner testimonials. Any team planning significant migration should test consistency, speech quality, and multi-turn editing cost savings on a pilot batch of their own content before committing production volume.

Watch for independent benchmarks to emerge over coming months and monitor whether Provisioned Throughput becomes necessary to meet production SLAs. Track whether the 15-second limit or documented gaps in sports action and complex physics affect your specific use case or become easier to work around as the model evolves.

For teams already using the Gemini API or evaluating Google Cloud for video generation, the pricing and clip-length improvements merit evaluation. The multi-turn editing workflow could represent the larger unlock for cost control and production velocity if it delivers on its promised efficiency.

Related reading

• Gemini 4 Carbon vs GPT-6.1 Sol: Pricing and Performance
• StreamVibe: Google’s video agent for timely help
• Gemini 4 Carbon: Google’s top-ranked model, defenders first