Veo 3 in 2026: Pricing, Veo 3.1 Tiers, and How to Use It

Veo 3 is Google DeepMind's text-to-video model, producing clips of about 8 seconds with native audio at $0.40 per second, while Veo 3 Fast costs $0.15 per second. By 2026, the newer Veo 3.1 adds Lite, Fast, and Standard tiers, putting the cost of an 8-second clip anywhere from about $0.40 to $4.80 depending on the tier and resolution you choose.

• Veo 3 = $0.40/second and Veo 3 Fast = $0.15/second through the Gemini API (Source: Google Developers Blog, 2025)
• Veo 3.1 Lite starts at $0.05/second (720p), the cheapest way to draft video with Google's models in 2026 (Source: Veo3AI.io, 2026)
• An 8-second clip costs $0.40 (Lite 720p) to $4.80 (Standard 4K), a 12x spread depending on tier and resolution
• In Thailand, Google AI Pro costs THB 750/month and Google AI Ultra costs THB 7,500/month (Source: Tech Offside, 2025)
• Creators who want steadier pricing can compare Veo 3 with an all-in-one platform like Imgentic, which offers VEO3.1 (including Fast and First-Last Frame) alongside Seedance 2.5 and Seedream on one credit balance

What Is Veo 3, and What's Different in 2026?

Veo 3 is Google DeepMind's text-to-video model, generating clips of about 8 seconds with native audio. It launched in May 2025 (Source: Wikipedia). Today Veo 3 sits alongside the newer Veo 3.1, which adds Lite, Fast, and Standard tiers and resolution up to 4K, supporting both text-to-video and image-to-video.

The production model veo-3.0-generate-001 reached general availability (GA) on July 29, 2025, according to Google Cloud's documentation (Source: Gemini Enterprise Agent Platform Docs, 2025). That date matters for teams building on Google Cloud Vertex AI, because it marks the point where Veo 3 moved from preview to a billable production product.

For creators, the change is straightforward. Earlier versions produced silent video with no sound, but Veo 3 generates picture and audio in a single pass. A talking-head clip or a street scene arrives with dialogue and ambient sound already attached.

Text-to-Video, Image-to-Video, and Native Audio

Text-to-video in Veo 3 means typing a text prompt and getting a finished clip back, while image-to-video means uploading a still, such as a product photo, and letting Veo 3 animate it while preserving the original composition. Both modes produce clips of roughly 8 seconds per generation and can include native audio.

Native audio generation is Veo 3's ability to create sound and dialogue synced directly to the video, without a separate audio pass. In practice, it removes an entire step: no round trip to audio software, no lip-sync fixes, no hunting through sound-effect libraries.

Tip: if your prompt includes dialogue, put the line in quotation marks and describe the speaker's tone of voice. In the cases we see most often, this produces tighter lip-sync than a loose instruction like "a man speaks."

How Veo 3 Differs from Veo 3.1

Veo 3.1 is the 2026 version of Google's video model. Its biggest difference from Veo 3 is pricing granularity. Veo 3 uses a single rate of $0.40/second (or $0.15 on Veo 3 Fast), while Veo 3.1 splits into Lite, Fast, and Standard, each with 720p and 1080p options and 4K on some tiers (Source: Veo3AI.io, 2026).

Output quality between the two versions is something you should test yourself before committing your budget.

8-Second Length and a 3-Clip Daily Quota

Veo 3 clips run about 8 seconds each, and Thai users on the Pantip forum reported a quota of 3 clips per day on consumer plans in 2025 (Source: Pantip, 2025). These two limits shape how you plan work. An 8-second maximum means a 60-second ad needs at least 8 generations, and a daily quota means the project stretches over several days unless you move to paid API billing.

Caution: quota numbers on consumer plans can change without notice. Treat the 3-clip figure as a 2025 community report, not a guaranteed entitlement in 2026.

How Much Does Veo 3 Cost per Second in 2026?

Veo 3 costs $0.40 per second and Veo 3 Fast costs $0.15 per second (Source: Google Developers Blog, 2025). In the 2026 Veo 3.1 lineup, the Lite tier starts at $0.05/second (720p), Fast runs $0.10–$0.30/second, and Standard runs $0.40–$0.60/second, putting a single 8-second clip anywhere from about $0.40 to $4.80 (Source: Veo3AI.io, 2026).

Per-second pricing is a billing model where you pay for every second of video you generate, so an 8-second clip costs the chosen tier's per-second rate multiplied by 8. That multiplication is exactly where budgets break, because a small-looking per-second number hides a much larger per-clip number.

Per-Second Pricing by Tier and Resolution

Google Veo 3 cost depends on two choices: the model tier and the output resolution. The table below puts every published rate for Veo 3, Veo 3 Fast, and all three Veo 3.1 tiers in one place, something no top-ranking page has yet done in full.

All figures are in US dollars. Sources: Veo 3 and Veo 3 Fast rates from the Google Developers Blog (2025); Veo 3.1 Lite, Fast, and Standard rates from Veo3AI.io (2026). Google adjusts pricing periodically, so check the latest rates on the Gemini API pricing page before you run large batches.

The Real Cost of One 8-Second Clip

One 8-second Veo 3 clip costs $3.20 on the full model and $1.20 on Veo 3 Fast. On Veo 3.1, the same 8-second clip ranges from $0.40 on Lite 720p to $4.80 on Standard 4K (Source: Veo3AI.io, 2026). Multiplying the per-second rate by 8 is the only honest way to compare tiers.

Bottom line: drafting 10 ideas on Lite 720p costs $4.00, while rendering the same 10 on Standard 4K costs $48.00. The quality you need for final delivery is almost never the quality you need to test whether a prompt works.

Free vs. Paid Access to Veo 3

Veo 3 isn't fully free in 2026. You access it through per-second billing on the Gemini API, a paid Google AI subscription, or a credit-based platform such as Imgentic, which offers VEO3.1, VEO3.1 Fast and First-Last Frame, and every free trial is capped by a quota (Source: Tech Offside, 2025). People searching for a "free AI video generator" usually end up with a limited daily quota, not an unlimited tool.

The trade-off is straightforward. Free or quota-limited access is fine for practicing prompts, while the paid API is what you need for client work, because you control the tier, resolution, and volume without waiting for a daily quota reset. To see how an all-in-one platform handles this, compare it with the best AI video generation platform options for your budget.

How to Access Veo 3 in Thailand, and What It Costs in Baht

Veo 3 is available in Thailand through Google AI Pro at THB 750 per month and Google AI Ultra at THB 7,500 per month (Source: Tech Offside, 2025). Subscription plans bundle Veo access with a usage quota, which can be cheaper than per-second billing on the Gemini API for creators who produce a steady, moderate volume of short clips.

The other route is the Gemini API, billed per second in US dollars. It has no monthly minimum, so in a month when you generate little, you pay little. The subscription route has a fixed minimum but a predictable ceiling.

Google AI Pro vs. Google AI Ultra

Google AI Pro at THB 750/month is the entry plan for users in Thailand who want Veo 3 in Google's apps, while Google AI Ultra at THB 7,500/month is the high-volume tier at ten times the price (Source: Tech Offside, 2025). Each plan's Veo quota changes over time, so check current entitlements before you subscribe.

Subscription vs. Per-Second Billing on the Gemini API

Gemini API billing charges $0.05–$0.60 per second depending on the Veo tier, while the Thai subscription plans charge a flat THB 750 or THB 7,500 per month regardless of how many seconds you generate within the quota. To compare them, convert the subscription to dollars: THB 750 comes to roughly $21–$23, depending on the exchange rate on the day you pay.

• At Veo 3.1 Lite 720p ($0.40 per 8-second clip), $21–$23 buys roughly 52–57 clips on the API
• At full Veo 3 ($3.20 per 8-second clip), the same money buys only about 6–7 clips
• At Veo 3.1 Standard 4K ($4.80 per clip), it buys about 4 clips

The upshot: the API clearly wins on volume of cheap drafts, while a subscription wins only when its quota delivers more high-tier clips than the figures above. API pros: no cost in months you don't use it, and full control over tiers. Cons: billed in dollars, and it requires developer setup. Subscription pros: one predictable payment in baht, and no code. Cons: a quota ceiling you can't raise mid-month.

Which Route Is Cheaper for Your Monthly Volume?

The cheaper route for creators making fewer than about 50 draft-quality clips a month is the Gemini API on Veo 3.1 Lite, because the total stays below the THB 750 subscription. Once you need a lot of Standard or 4K clips every month, Google AI Ultra at THB 7,500 may be cheaper than the equivalent API spend, provided its quota covers your volume.

Tip: log your real clip count for one month before deciding. In the cases we see most often, freelancers overestimate how many final-quality clips they make and underestimate their drafts, which means the API is actually a better fit for them than a subscription.

Veo 3 vs. Veo 3.1: Which Should You Use?

Veo 3.1 is the newer version, with tiered pricing (Lite, Fast, Standard) and support for up to 4K, which gives you finer control over cost. The original Veo 3 uses a single rate of $0.40/second, with a Fast option at $0.15/second (Source: Google Developers Blog, 2025; Veo3AI.io, 2026). Choose Veo 3.1 Lite for cheap 720p drafts, and use Standard or 4K only for final delivery.

How the Pricing Models Differ

Veo 3 pricing has just two options: $0.40/second on the main model or $0.15 on Veo 3 Fast, with no discount by resolution. Veo 3.1 pricing is a ladder, with Lite at $0.05–$0.08, Fast at $0.10–$0.30, and Standard at $0.40–$0.60 per second (Source: Veo3AI.io, 2026). That ladder rewards people who separate drafting from final rendering.

Resolution and Quality Options

Veo 3.1 supports 720p, 1080p, and 4K output (Source: Google AI for Developers), while Veo 3 tops out at 1080p under a single rate. 4K generation on Veo 3.1 is available on the Fast tier at $0.30/second and on Standard at $0.60/second, so 4K is never cheap, but it's now genuinely possible without upscaling.

Realism across tiers is something to test with your own footage before you promise a client a particular look.

Which Tier Fits Which Job

The best Veo tier for testing prompts is Veo 3.1 Lite 720p at $0.40 per clip, because you can iterate ten times for the price of one Standard render. Veo 3.1 Fast 1080p at $0.96 per clip suits social posts, and Veo 3.1 Standard 4K at $4.80 per clip should be reserved for hero shots and broadcast work.

• Lite 720p: storyboard checks, prompt iteration, internal previews
• Fast 1080p: Reels, TikTok, YouTube Shorts, cuts for client review
• Standard 1080p/4K: final ads, cinematic short films, big-screen work
• Original Veo 3: existing pipelines already tuned to veo-3.0-generate-001 that you don't want to re-test

Common Mistakes and Cautions When Using Veo 3

The most common Veo 3 mistake is underestimating how costs creep up. Rendering everything in 4K on the Standard tier pushes an 8-second clip to $4.80, 12 times the $0.40 cost of a draft on Veo 3.1 Lite (Source: Veo3AI.io, 2026). Creators also forget about the 8-second clip limit and the daily quota, then burn their budget on re-renders instead of drafting cheaply first.

These are the mistakes we see most often when reviewing creators' budgets, with a fix for each one.

• Rendering drafts in 4K. Ten 4K Standard drafts cost $48.00, while ten Lite 720p drafts cost $4.00. Lock the prompt on Lite first, then render in 4K once.
• Forgetting the 8-second ceiling. A 60-second ad is at least 8 generations, and often 12–15 after retakes. Budget per scene, not per video.
• Relying on a 3-clip daily quota for client deadlines. The 2025 Pantip report of 3 clips per day (Source: Pantip, 2025) means consumer plans can't be rushed. Move deadline work to the paid API.
• Skipping the cheap tier entirely. Testing a new prompt style on Standard at $3.20 a try burns money on failures that Lite would have revealed for $0.40.
• Not tracking the exchange rate. The API bills in dollars and subscriptions bill in baht. In a month when the baht weakens, API costs rise by several percent even though your volume hasn't changed.

Overpaying for 4K on Drafts

4K on Veo 3.1 Standard costs $0.60/second, 12 times the Lite 720p rate of $0.05/second (Source: Veo3AI.io, 2026). On a budget of 100 clips a month, that gap is $480 versus $40. Early drafts almost never need 4K, because you're judging motion, framing, and prompt adherence, none of which change with pixel count.

Caution: some pipelines default to the highest resolution. Check the resolution parameter on every API call or app setting before you hit generate.

Ignoring the 8-Second Clip Limit

Every Veo 3 generation returns a clip of about 8 seconds, so anything longer is a multi-clip job with multiplied costs. A 30-second product video is at least 4 clips, and continuity between cuts often forces you to regenerate adjacent scenes. In practice, budget for 6–8 clips. Planning by scene count keeps your estimates honest.

Not Testing Prompts on a Cheap Tier First

Testing on Veo 3.1 Lite first is the biggest money-saver in 2026. Iterate on a prompt at $0.40 a try until the composition, camera movement, and dialogue land, then pay $3.20 or $4.80 once for the final render. Creators who reverse this order often pay 5–10 times more per finished clip.

How to Create AI Video or an AI Film from Text with Veo 3

To create AI video from text with Veo 3, you write a detailed prompt, choose a tier and resolution, then generate an 8-second clip with native audio through the Gemini API (Source: Google AI for Developers) or a Google AI plan. For longer AI films, you stitch multiple clips together or use an intelligent platform that sequences scenes automatically from start to finish.

Searches like "make a movie with AI," "AI short film maker," and "AI commercial video" all describe the same workflow: generate short clips, then assemble them into a longer story. The steps below work for both a 15-second ad and a 3-minute short film.

Writing Text-to-Video Prompts That Work

An effective Veo 3 prompt describes the character, action, setting, camera, lighting, and sound, in that order, in one or two clear sentences. Vague prompts produce generic footage and waste generations, while specific prompts give Veo 3 enough structure to nail the shot on the first or second try.

1. Define the character clearly: "a 30-year-old Thai barista in a green apron," not "a person"
2. Give one action that fits within 8 seconds: "pours latte art into a white cup"
3. Describe the setting and time of day: "a small Bangkok café, morning light through the window"
4. Specify the camera: "slow push-in, shallow depth of field, 35mm lens look"
5. Add the sound you want: "soft hiss of the espresso machine, light jazz in the background, no dialogue"
6. Run the prompt on Veo 3.1 Lite 720p first, then change one variable per retry

Generating and Downloading Clips

Generating a Veo 3 clip follows the same few steps in principle, whether you use the Gemini API or a consumer plan: submit the prompt, choose the tier and resolution, wait for the render, review it, then download. If you're just starting out, read the guide to creating AI video from text first, then come back here to plan your costs.

1. Sign in to Google AI Studio or a Google AI Pro/Ultra account, or set up a Gemini API key
2. Choose the Veo model and tier (Lite, Fast, or Standard), and set the resolution to 720p for the first pass
3. Paste in the prompt, or upload a reference image for image-to-video mode
4. Click generate and wait for the 8-second clip with native audio to finish rendering
5. Review picture and sound together, and note what needs fixing before you pay for a higher tier
6. Run the approved prompt on Standard 1080p or 4K for final delivery only, then download

Stitching Short Clips into a Full AI Film

A full AI film made with Veo 3 is many 8-second clips edited together, with characters and settings kept consistent across prompts. A 2-minute short film comes to about 15 clips plus retakes. At $0.40 per Lite draft and $3.20 per final Standard clip, a realistic budget is $60–$100 before music and editing.

You can stitch the clips yourself, but that means writing 15 prompts, managing continuity by hand, and balancing audio levels between clips. An intelligent platform lifts that burden by planning scenes, generating them in order, and assembling the timeline for you. If that fits your workflow, see how to make an AI film end to end.

What Are the Alternatives to Veo 3, and Which All-in-One Platforms Suit Creators?

Beyond Veo 3, creators can use intelligent all-in-one platforms that bring multiple models into a single workflow. Imgentic (imgentic.ai) wraps models like Seedance 2.5 and Seedream into one workspace for both images and video, and also offers VEO3.1, VEO3.1 Fast, First-Last Frame and Reference-to-video on the same credits.