Google Veo 3
Google's Veo 3.1 video model: up to 4K clips with native audio via Gemini API and Flow
Veo 3.1 generates 4-, 6- or 8-second video from text or image prompts, with dialogue, sound effects and ambience created in the same pass. Standard reaches 4K on 8-second clips, and Fast and Lite tiers cut the per-second API price. It is reached through the Gemini API, Google Flow and the Gemini app, with Vertex AI as the enterprise route.
- 54 creators using
- 5 news items tracked
- 2025 first launched
- Paid pricing
- Easy learning curve
What Google Veo 3 can do
- Infinite canvas Not confirmed
- AI agent Not confirmed
- Node / flow builder Not confirmed
- Public API Yes
- MCP server Not confirmed
8 of 26 capabilities confirmed
| Capability | Supported | Details |
|---|---|---|
| Text to image | No | |
| Image editing | No | |
| Text to video | Yes | |
| Image to video | Yes | Image inputs; up to 3 reference images on 8-second clips |
| Video to video | Partial | Extension of Veo-generated clips only (720p, up to 20 extensions) |
| Native audio in video | Yes | Dialogue, sound effects and ambience generated with the video, always on |
| Lip-sync | Not confirmed | |
| Voice / TTS | No | |
| Music & SFX | No | |
| Upscaling | Not confirmed | |
| Character consistency | Partial | Up to 3 reference images, 8-second clips only |
| Camera control | Partial | Camera moves are prompt-directed; no dedicated camera-control UI verified |
| Custom training | No | |
| Third-party models | No |
| Capability | Supported | Details |
|---|---|---|
| Infinite canvas | Not confirmed | |
| Node / flow builder | Not confirmed | |
| AI agent | Not confirmed | |
| Chat assistant | Partial | Prompting through the Gemini app chat; conversational editing of existing clips is an Omni Flash feature, not Veo |
| Team collaboration | Not confirmed |
| Capability | Supported | Details |
|---|---|---|
| Public API | Yes | Gemini API (veo-3.1-generate-preview and Fast/Lite variants); also Vertex AI |
| MCP server | Not confirmed | |
| Mobile app | Not confirmed | |
| Desktop / local | No | |
| Open source | No |
| Capability | Supported | Details |
|---|---|---|
| Free tier | No | Gemini API pricing page: free tier not available for Veo. Google's subscriptions page lists no Veo video allowance on the free plan |
| Commercial use | Not confirmed |
Yes Partial No Not confirmed by the vendor
What makes Google Veo 3 different
- 01
Native audio is generated with the picture (dialogue, SFX, ambience) and is included in the listed per-second price, not a separate charge.
- 02
4K output on 8-second clips, where most competitors top out at 1080p in the same price band (Veo's own 4K tier costs $0.60 per second).
- 03
Clip extension: a Veo clip can be extended by 7 seconds up to 20 times, reaching 148 seconds, with the extension run at 720p.
- 04
Transparent API pricing per second across three tiers (Lite $0.05, Fast $0.10, Standard $0.40 at 720p), so shot budgets can be calculated before rendering.
- 05
Same model family reachable from the Gemini API, Gemini app and Flow, so a shot tested in Flow can be scripted in the API.
Latest from Google Veo 3
- 2d ago Veo 3 integrates Gemini 4 Carbon for higher‑fidelity video generation Veo 3 adds DeepMind’s Gemini 4 Carbon model, boosting resolution and temporal consistency for creators.
- 3d ago Genie 4 Introduces Real-Time World Simulation and Interactive Control Google DeepMind introduced Genie 4, a world model that generates interactive, physics-grounded environments in real time to expand beyond static video generation.
- 1w ago Google DeepMind maps global forest types at 10-meter resolution Google DeepMind's new AI model categorizes global land into 12 distinct forest types at a 10-meter resolution to improve ecological tracking.
- 2w ago Antigravity model accelerates discovery of new antimicrobial peptides Google DeepMind’s Antigravity model identifies potent antimicrobial peptides to help researchers develop new treatments for antibiotic-resistant bacteria.
- 2w ago On-device Gemma models detect conversational scams locally Google DeepMind and Singapore’s GovTech deploy lightweight Gemma models for real-time, private scam detection on mobile devices.
About Google Veo 3
Veo 3.1 is the current model in Google DeepMind's Veo family. Compared with Veo 3, the verified changes are the 3.1 model IDs, reference-image inputs (up to 3 images, 8-second clips only), and a video extension feature that adds 7 seconds per step, up to 20 steps, for clips up to 148 seconds. Native audio is always generated, and 16:9 and 9:16 are the aspect ratios. Standard, Fast and Lite tiers exist for the Gemini API, with Lite dropping 4K.
The model is aimed at short narrative shots, previs, and B-roll where a single prompt needs to produce picture and sound together. Clips are capped at 8 seconds per generation, so longer sequences need extensions or cutting several generations together. Extension only works on Veo-generated footage and only at 720p.
Google has moved its video line forward. Gemini Omni Flash, launched at I/O on May 19, 2026, adds conversational editing of existing clips and is what the Gemini API docs now name as the default video model. Veo 3.1 remains available, so the choice is between Omni's editing and multimodal input and Veo's per-second API pricing and 4K output.
For filmmakers, the practical route is the Gemini API or Flow. Flow uses subscription credits, and the API is billed per generated second with no free tier.
What's new in Google Veo 3
-
Gemini Omni Flash launched; named default video model in Gemini API docs
Omni Flash adds conversational editing and multimodal input and produces clips up to about 10 seconds with audio, alongside Veo 3.1.
Source: ai.google.dev ↗ -
Veo 3.1 Lite added to the Gemini API
Lowest-cost Veo tier at $0.05 per second at 720p and $0.08 at 1080p, with no 4K output.
Source: ai.google.dev ↗
Models available in Google Veo 3
-
Veo 3.1 Latest $$$ Google Standard tier: 720p/1080p/4K, 4/6/8s, 16:9 and 9:16, native audio Also: Veo 3.1 Fast · Veo 3.1 Lite
-
$ · $$ · $$$ = relative cost against other models of the same type, on fal.ai list prices. Cost inside Google Veo 3 depends on its own plans and credits.
Key Features
- Text-to-video and image-to-video at 4, 6 or 8 seconds
- Native audio: dialogue, sound effects and ambience generated with the picture
- 720p default, 1080p and 4K on 8-second clips (Standard and Fast)
- 16:9 and 9:16 aspect ratios
- Up to 3 reference images (8-second clips only)
- Video extension: +7s per step, up to 20 steps, up to 148 seconds total
- Frame-specific generation (first and last frame direction)
- Three API tiers (Standard, Fast, Lite) with per-second billing
- Gemini API and Vertex AI access for developers and studios
- Gemini app and Google Flow for consumer and prosumer use
When to reach for it — and when to skip
Reach for it when…
- Native audio in the same pass, with no separate dubbing or foley step for simple shots
- 4K output on 8-second clips, with per-second pricing that is published and easy to budget
- Extension lets one prompt grow into a 2-minute continuous shot
- Cheap Lite and Fast tiers make exploration and previs affordable
Skip it when…
- 8-second maximum per generation; longer shots need extensions or stitching
- Extension only works at 720p and only on Veo-generated footage
- Most of the API models are still labelled preview, so IDs and pricing can change
- Google is positioning Gemini Omni as the successor, so the Veo roadmap is less clear than it was
Best For
✓ Ideal for
- Previs and animatics for narrative scenes where audio matters from the first pass
- Short commercial or music-video B-roll at 4K on 8-second shots
- Exploring shot options cheaply with Veo 3.1 Lite before paying for Standard
- Developers and studios scripting batches of clips through the Gemini API or Vertex AI
✗ Not built for
- Music generation (Veo produces sound effects and dialogue, not scored music)
- Clips longer than about 2.5 minutes, or per-shot camera rigs with exact moves
Working Tips from Filmmakers Using Google Veo 3
- 01 Budget by the second: a 4K 8-second Standard clip is $4.80 on the API, while the same length at Lite 720p is $0.40. Test prompts on Lite and render keepers on Standard
- 02 Reference images only work on 8-second clips, so set duration to 8 before attaching them
- 03 Extend only Veo-generated footage at 720p. Plan the final resolution before extending because extension output stays at 720p
Google Veo 3 pricing in 2026
| Plan | Price | What you get |
|---|---|---|
| Gemini API Veo 3.1 Lite | $0.05/s (720p), $0.08/s (1080p) usage |
|
| Gemini API Veo 3.1 Fast | $0.10/s (720p), $0.12/s (1080p), $0.30/s (4K) usage |
|
| Gemini API Veo 3.1 Standard | $0.40/s (720p and 1080p), $0.60/s (4K) usage |
|
| Google AI Plus | EUR 4.99 monthly |
|
| Google AI Pro | EUR 21.99 monthly |
|
| Google AI Ultra | EUR 99.99 (5x limits) or EUR 219.99 (20x limits) monthly |
|
Prices checked against Google Veo 3's pricing page on 9 Oct 2026. Vendors change plans often, so confirm before you buy.
Google Veo 3 vs Runway vs Kling AI vs Higgsfield
| Google Veo 3 | Runway | Kling AI | Higgsfield | |
|---|---|---|---|---|
| Price | Gemini API: $0.05 to $0.60 per generated second depending on tier and resolution. Consumer access via Google AI Plus from EUR 4.99/mo (200 Flow credits). | Free (125 one-time credits); Standard $15/mo ($12/mo billed annually) | See the vendor's pricing page | Free tier; paid plans from €19/mo billed annually (prices exclude VAT) |
| Infinite canvas | Not confirmed | Not confirmed | Partial | Partial |
| Node / flow builder | Not confirmed | Yes | Not confirmed | Not confirmed |
| AI agent | Not confirmed | Yes | Not confirmed | Yes |
| Chat assistant | Partial | Yes | Not confirmed | Yes |
| Team collaboration | Not confirmed | Yes | Not confirmed | Yes |
| Text to image | No | Yes | Yes | Yes |
| Image editing | No | Yes | Not confirmed | Yes |
| Text to video | Yes | Yes | Yes | Yes |
| Image to video | Yes | Yes | Yes | Yes |
| Video to video | Partial | Yes | Yes | Yes |
| Native audio in video | Yes | Partial | Yes | Yes |
| Lip-sync | Not confirmed | Not confirmed | Yes | Not confirmed |
| Voice / TTS | No | Yes | Partial | Yes |
| Music & SFX | No | Yes | Not confirmed | Yes |
| Upscaling | Not confirmed | Yes | Not confirmed | Yes |
| Character consistency | Partial | Partial | Yes | Yes |
| Camera control | Partial | Not confirmed | Yes | Yes |
| Custom training | No | Partial | Not confirmed | Not confirmed |
| Third-party models | No | Yes | No | Yes |
| Public API | Yes | Yes | Yes | Yes |
| MCP server | Not confirmed | Yes | Not confirmed | Yes |
| Mobile app | Not confirmed | Yes | Not confirmed | Partial |
| Free tier | No | Yes | Yes | Yes |
| Commercial use | Not confirmed | Not confirmed | Yes | Not confirmed |
Each column comes from that tool's own researched page. "Not confirmed" means the vendor doesn't say.
Open the full Google Veo 3 vs Runway vs Kling AI vs Higgsfield comparison →Use Cases
Tags
Creators Using Google Veo 3
AI filmmaker converting ideas to story that sells and move
Ai filmmaker, commercial ad luxury, branding design and vision
Ai Filmmaker • Cinematic Storyteller • Short Film Creator
Sources
Facts on this page were checked against Google Veo 3's own pricing, product and changelog pages on , then cross-checked with independent coverage where vendor pages were silent. Anything we couldn't confirm is marked "Not confirmed" rather than guessed.
- Gemini API pricing (Veo section) ai.google.dev
- Veo 3.1 Gemini API docs (durations, resolutions, extension) ai.google.dev
- Gemini API video generation overview (Omni Flash as default) ai.google.dev
- Google DeepMind Veo page deepmind.google
- Gemini subscriptions and Flow credits gemini.google
- Runway pricing (comparison column) runway.com
Spotted something out of date? Tell us and we'll re-check it.
Discussion
No comments yet — be the first.