Description
Veo 3.1 is Google DeepMind’s current text-to-video and image-to-video model, the successor to the original Veo 3 released in May 2025. It generates cinematic video clips with native audio — dialogue, ambient sound, and music generated and synchronized alongside the visuals rather than added afterward — and supports resolutions up to 4K on the developer API. Google describes Veo as designed to empower filmmakers and storytellers, built with real-world physics and improved prompt adherence over prior versions. Consumer access runs through Google Flow (Google’s dedicated AI filmmaking tool), the Gemini app, and Google Photos, bundled into Google’s AI subscription tiers alongside Gemini chat and storage rather than sold as a standalone product. Developers can also access Veo 3.1 directly, billed per second of generated video through the Gemini API or Vertex AI, separately from the consumer subscription tiers. Every video generated with Veo is marked with SynthID, Google’s watermarking technology for identifying AI-generated content.
Key Features
- Native audio generation — produces dialogue, ambient sound, and music synchronized to the video, generated alongside the visuals rather than layered on afterward.
- Up to 4K resolution — the developer API supports 720p, 1080p, and 4K output, depending on model tier.
- Three model tiers — Standard, Fast, and Lite, trading resolution, audio quality, and speed against per-second cost.
- Real-world physics — models motion, lighting behavior, and object interaction for grounded, believable movement.
- SynthID watermarking — every generated video carries Google’s invisible watermarking technology for AI-content identification, applied automatically.
- Multiple access points — available through Google Flow, the Gemini app, Google Photos, and developer access via the Gemini API and Vertex AI.
How It Works
A user enters a text prompt or uploads a reference image inside Google Flow, the Gemini app, or Google Photos, and Veo 3.1 generates a short clip with synchronized native audio built directly into the render. Developers instead call the model directly through the Gemini API or Vertex AI, selecting the Standard, Fast, or Lite variant depending on how much resolution and audio fidelity a project needs against its per-second budget. Consumer usage draws from a monthly Google Flow credit allowance tied to a Google AI subscription tier, while API usage is billed independently, per second of successfully generated video — failed generations aren’t charged. Every output, consumer or API, is marked with SynthID, Google’s watermarking system for identifying AI-generated content, and passes through Google’s safety evaluations before release.
Technical Architecture & Overview
- Core Engine: Google DeepMind’s Veo 3.1 model family (Standard, Fast, Lite variants), successor to Veo 3 (May 2025).
- Deployment: Google Flow, the Gemini app, and Google Photos for consumers; Gemini API and Vertex AI for developers.
- API Surface: Per-second billing through the Gemini API/Vertex AI, separate from the Google AI subscription tiers; documented on Google’s official Gemini API pricing page.
- Known Limits: No standalone consumer price — bundled only into Google AI subscription tiers; original Veo 3 and Veo 2 model IDs were deprecated and shut down June 30, 2026.
Pros & Cons
| Pros | Cons |
|---|---|
| Native audio is generated and synchronized alongside the video itself, not layered on as a separate step. | Veo 3.1 has no standalone price of its own; it’s only available bundled into a Google AI subscription tier or billed separately through the API. |
| The developer API supports resolution up to 4K, ahead of most competing video-generation APIs. | Pricing is spread across several Google AI tiers and a separate per-second API, making total cost harder to estimate upfront. |
| Three model tiers (Standard, Fast, Lite) let a project trade quality against per-second API cost. | The original Veo 3 and Veo 2 model IDs were deprecated and shut down by Google on June 30, 2026, requiring a migration to Veo 3.1. |
| Deep integration across Google Flow, the Gemini app, Google Photos, and the Gemini API gives several different access paths. | Standard-tier API pricing at $0.40 per second is the most expensive rate among its major competitors’ base models. |
| Every output carries Google’s SynthID watermark, giving built-in provenance tracking for generated content. | 4K output is only available through the developer API, not through the consumer Google Flow or Gemini app experience. |
Pricing
| Plan | Price |
|---|---|
| Google AI Plus | $7.99/month (200 Flow credits/month) |
| Google AI Pro | $19.99/month (1,000 Flow credits/month) |
| Google AI Ultra | $100–$200/month (10,000–25,000 Flow credits/month) |
https://one.google.com/about/google-ai-plans
API / Developer Pricing
Veo 3.1 is also billed per second through the Gemini API, separately from the Google AI subscription tiers above, confirmed on Google’s official Gemini API pricing page:
| Model Tier | API Price (with audio) |
|---|---|
| Veo 3.1 Lite | $0.05/sec (720p), $0.08/sec (1080p) |
| Veo 3.1 Fast | $0.10/sec (720p), $0.12/sec (1080p), $0.30/sec (4K) |
| Veo 3.1 Standard | $0.40/sec (720p and 1080p), $0.60/sec (4K) |
Platform Availability
Google Flow | Gemini app | Google Photos | Gemini API | Vertex AI
Best For
Google Workspace and Gemini users wanting native-audio video inside familiar Google apps | Developers needing resolution-flexible, per-second billed video generation | Storytellers and filmmakers wanting cinematic 4K output via the API
Similar Tools in This Directory
Also worth comparing: Kling AI, character-consistent video with native audio and a separate prepaid API | Sora AI, OpenAI’s text-to-video model now bundled into ChatGPT Plus and Pro | Runway, standalone video platform with its own Gen-4.5 model plus multi-model credit access
Visit Official Website
Frequently Asked Questions
Does Veo 3.1 have a free plan?
Not on its own. Veo 3.1 has no standalone price and is only available bundled into a paid Google AI subscription tier, starting at Google AI Plus ($7.99/month), according to Google’s official AI plans page.
How much does Veo 3.1 cost?
Google AI subscription tiers with Flow credits for Veo 3.1 run from $7.99/month (Plus) to $200/month (Ultra 20x). Developers can also access Veo 3.1 directly through the Gemini API, billed from $0.05/second (Lite) to $0.40/second (Standard), according to Google’s official Gemini API pricing page.
Does Veo 3.1 generate audio?
Yes. Veo 3.1 generates dialogue, ambient sound, and music synchronized to the video, built in alongside the visuals rather than added afterward, according to Google DeepMind’s official model page.
What resolution does Veo 3.1 support?
Up to 4K through the developer API, with 720p and 1080p available across all three model tiers (Standard, Fast, Lite), according to Google’s official Gemini API pricing page.
What happened to the original Veo 3?
Google deprecated and shut down the original Veo 3 and Veo 2 model IDs on June 30, 2026, migrating developers to Veo 3.1, according to Google’s official Gemini API documentation.
