Unlimited Open Source Models

Get Plan
Skip to main content
Imagen

Cheapest AI Video API for Developers

Plans from $21/mo. Self-hosted video clips start at $0.05, and eight self-hosted video models are unlimited on $149/mo. Partner models such as Kling and Veo are priced per second.

Get API keySee pricing
  • Self-hosted clips from $0.05
  • 8 video models unlimited on $149/mo
  • Plans from $21/mo

What Actually Drives AI Video Cost

The Model Sets the Price, Not the Platform

Video is the most expensive thing to generate with AI, and most "cheapest video API" comparisons hide where the money goes. The cost of a clip is driven by the model, the resolution and the duration — not by which platform you call it through. A 10-second 1080p clip from a premium model costs real GPU time no matter who bills you for it.

ModelsLab is honest about the split. Eight open-weight video models run on our own GPUs. On Basic and Standard a clip costs $0.05 to $0.075 on most of them, charged from your plan allowance, and on Open Source Unlimited they have no per-clip charge. Partner models — Kling, Veo, Seedance, Hailuo, Wan 2.7 and others — are priced per second: from the usage included in Basic and Standard first, then from wallet balance, and from wallet balance on Open Source Unlimited.

Why That Makes the $21 Plan Genuinely Cheap

The Basic plan is $21/month, and that $21 is your usage allowance. At $0.075 a Wan 2.2 clip, it covers about 280 clips, so you can prototype and iterate on open-weight models without touching wallet balance. Here is what a plan reaches:

  • Text-to-video and image-to-video on one REST endpoint
  • Video-to-video restyling and scene creation
  • 8 open-weight video models covered by your plan
  • Kling, Veo, Seedance and other partner models priced per second
  • Prompt, duration, aspect ratio and motion controls
  • Async job submission with a fetch endpoint for results
  • Standard plan at $47/month with 10 parallel generations
  • One API key shared with image, speech and LLM

What Each Plan Costs

Each plan is a dollar allowance, not a call count. What a clip costs depends on whether the model runs on our GPUs or a partner's.

PlanPrice / MonthSelf-hosted videoParallel generations
Basic (best entry point)$21Per clip from the $21 allowance5
Standard$47Per clip from the $47 allowance10
Open Source Unlimited$149Unlimited on 8 self-hosted models15

ModelsLab plan pricing, verified against the live plans table on 2026-09-23. One subscription covers video, image, speech and LLM APIs. Open Source Unlimited applies to the 8 open-weight video endpoints ModelsLab self-hosts. Partner video models (Kling, Veo, Seedance, Hailuo, Wan 2.7 and others) are priced per second and are paid per call from the usage included in Basic ($21) and Standard ($47), then from wallet balance; on Open Source Unlimited they are billed from wallet balance, because unlimited does not include them.

Self-hosted Video Models: Price per Clip

These eight models run on our own GPUs. On Basic and Standard each clip is charged from the plan allowance; on Open Source Unlimited there is no per-clip charge.

ModelBasic and StandardOpen Source Unlimited ($149)
SVD$0.05 a clipUnlimited on $149
CogVideoX$0.05 a clipUnlimited on $149
Wan 2.2$0.075 a clipUnlimited on $149
LTX-2.3$0.075 a clipUnlimited on $149
Wan 2.1$0.375 a 5-second clipUnlimited on $149
MiniMax H3 text-to-video$0.05 per secondUnlimited on $149
MiniMax H3 reference-to-video$0.05 per secondUnlimited on $149
MiniMax H3 start-end-frame$0.05 per secondUnlimited on $149

Partner models (Kling, Veo, Seedance, Hailuo, Wan 2.7) are priced per second. On Basic and Standard they draw on the plan allowance, then the wallet; on Open Source Unlimited they bill from the wallet.

Generate Video in One Request

Text-to-video and image-to-video on the same key.

Text to video on an open-weight model (Python)

Python
1import requests
2

3response = requests.post(
4    "https://modelslab.com/api/v7/video-fusion/text-to-video",
5    headers={
6        "key": "YOUR_MODELSLAB_API_KEY",
7        "Content-Type": "application/json",
8    },
9    json={
10        # wan2.2 and ltx-2.3 run on ModelsLab GPUs — covered by your plan
11        "model_id": "wan2.2",
12        "prompt": "a paper boat drifting down a rain-soaked gutter, macro shot",
13        "duration": 5,
14        "aspect_ratio": "16:9",
15    },
16)
17

18# Long renders return an id — poll the fetch endpoint for the result
19print(response.json())

Image to video (Python)

Python
1import requests
2

3response = requests.post(
4    "https://modelslab.com/api/v7/video-fusion/image-to-video",
5    headers={
6        "key": "YOUR_MODELSLAB_API_KEY",
7        "Content-Type": "application/json",
8    },
9    json={
10        "model_id": "wan2.2",
11        "prompt": "slow orbit around the product, soft studio light",
12        # init_image is a list, even for a single frame
13        "init_image": ["https://example.com/product-shot.jpg"],
14        "duration": 5,
15    },
16)
17

18print(response.json())

Poll for a finished render (JavaScript)

JavaScript
1const headers = {
2  'Content-Type': 'application/json',
3  key: 'YOUR_MODELSLAB_API_KEY',
4};
5

6// The fetch endpoint takes the request id as the last path segment.
7const fetchUrl = (id) =>
8  'https://modelslab.com/api/v7/video-fusion/fetch/REQUEST_ID'.replace(
9    'REQUEST_ID',
10    id,
11  );
12

13const submit = await fetch(
14  'https://modelslab.com/api/v7/video-fusion/text-to-video',
15  {
16    method: 'POST',
17    headers,
18    body: JSON.stringify({
19      model_id: 'wan2.2',
20      prompt: 'timelapse of a city square at dusk',
21      duration: 5,
22      aspect_ratio: '16:9',
23    }),
24  },
25);
26

27const { id, output } = await submit.json();
28

29// Already done? Use it. Otherwise poll the fetch endpoint by id.
30let video = output?.[0];
31while (!video) {
32  const poll = await fetch(fetchUrl(id), { method: 'POST', headers });
33  video = (await poll.json()).output?.[0];
34}
35

36console.log(video);

Ship Video in 3 Steps

Prototype on open-weight models, spend on partner models only where it counts.

STEP 01
STEP 01

Step 1: Sign Up

Create a ModelsLab account, subscribe to the $21/month Basic plan, and copy your API key from the dashboard.

STEP 02
STEP 02

Step 2: Prototype on Open-Weight Models

POST a prompt to the text-to-video endpoint with an open-weight model_id. These run on ModelsLab GPUs and are charged per clip from your plan allowance (Wan 2.2 is $0.075 a clip) — no wallet balance needed while you iterate.

STEP 03
STEP 03

Step 3: Add Premium Models Where Quality Pays

Switch model_id to Kling, Veo, Seedance or Minimax for hero shots. Those are priced per second (from the usage included in Basic or Standard, then wallet balance), so you spend premium budget only on the clips that ship.

Get API Key

Related video guides

Best AI Video API 2026

Full comparison beyond price — model quality, duration limits, and latency.

Text to Video API

Endpoint reference for generating video from a text prompt.

Image to Video API

Animate a still image into video through one REST call.

Lip Sync API

Match mouth movement to a new audio track for dubbing.

Seedance API

Seedance models on the same key, priced per second.

Why Unlimited Covers Self-Hosted Video Only

The $149/month Open Source Unlimited plan does say unlimited, and it is — for the eight open-weight video models ModelsLab self-hosts. It does not cover partner models such as Kling, Veo and Seedance, and we would rather say so up front than let you discover it after subscribing.

If your pipeline runs mostly on open-weight video, the unlimited plan removes per-clip cost entirely and is the cheapest way to run at volume. If it runs mostly on Kling or Veo, a metered plan plus wallet balance will cost less than paying for unlimited access you cannot use.

How to Actually Keep Video Costs Down

Generate drafts at low frame counts and short durations, then re-render only the take you keep at full settings. Most of the spend in a video pipeline goes on versions nobody ships.

Route by job rather than picking one model for everything. Bulk, background and draft renders belong on open-weight models covered by your plan; partner models are worth their per-second price for the handful of clips that face the customer. One endpoint and one key handle both, so routing is a change to model_id and nothing else.

Why Developers Start at $21

Key advantages that set us apart

Plans from $21/month, billed as a dollar allowance
Standard at $47/month with 10 parallel generations
Self-hosted clips from $0.05 on Basic and Standard
Partner models priced per second
Text-to-video, image-to-video and video-to-video on one endpoint
Async submission with a fetch endpoint for long renders
One API key across image, video, speech and LLM
Unlimited self-hosted video on 8 models at $149/month

Our Popular Use Cases

What teams build with low-cost video generation:

Animate existing product stills into short loops for listings and ads. Draft on open-weight models, then re-render the winners on a premium model.

Product Video at Scale

Related Video APIs

Your Data is Secure: GDPR Compliant AI Services

ModelsLab GDPR Compliance Certification Badge

GDPR Compliant

Start Creating Amazing AI Videos in Minutes

Choose plan as per your needs, cancel anytime.

100% refund policy on monthly & yearly plans — cancel anytime

Contact Sales
Best Value

Open Source Unlimited

Mission-Critical

$149/month

100% refund policy · cancel anytime

Unlimited Open Source Models
100% refund policy
24x7 Support
15 parallel generations
Access to all APIs
Unlimited generations on all open-source models
For mission critical workloads
Add Team Members
Priority GPU Clusters
Most Popular

Standard

Production

$47/month

100% refund policy · cancel anytime

Moderate Traffic
100% refund policy
Priority Developer Support
10 concurrent API requests
For Production workloads
API access to all models
Prototype

Basic

Prototype

$21/month

100% refund policy · cancel anytime

Moderate Traffic
100% refund policy
Developer Support via Discord/Email
5 concurrent API requests
API access to all models
Shared GPU

Start with the $21 plan

Self-hosted clips from $0.05. The same key works for image, video, speech and LLM APIs.

Get API key

Get Expert Support in Seconds

We're Here to Help.

Want to know more? You can email us anytime at support@modelslab.com

View Docs

ModelsLab plans start at $21/month (Basic), with Standard at $47/month. Open-weight video models ModelsLab self-hosts are paid from the usage included in your plan (Wan 2.2 is $0.075 a clip) and have no limit on the $149 Open Source Unlimited plan. Premium third-party models such as Kling and Veo are paid per call from the same included usage, then from wallet balance; on Open Source Unlimited they are billed from wallet balance.

No. The $149/month Open Source Unlimited plan gives unlimited generation on the 8 open-weight video models ModelsLab self-hosts. Partner video models — Kling, Veo, Runway, Seedance, Vidu, Tencent and others — are priced per second: on that plan from wallet balance, and on Basic and Standard from the usage included in the plan first, then from wallet balance.

The open-weight video models ModelsLab runs on its own GPUs, such as Wan 2.2, LTX-2.3 and SVD, are paid from the usage included in your plan and have no per-clip charge on the $149 plan. Partner models are listed with their own per-second price so the cost is visible before you send the request.

A clip is many frames plus the work of keeping them consistent, so one video uses far more GPU time than one image. That is why video pricing varies by model, resolution and duration, while a self-hosted image costs $0.0047 whatever the model.

ModelsLab is a paid service. Plans start at $21/month. Creating an account is free, but generating video requires an active plan.

Draft at short durations on open-weight models covered by your plan, then re-render only the take you are shipping at full settings on a premium model. Switching between them is a change to model_id — the endpoint and API key stay the same.

Yes. Text-to-video, image-to-video and video-to-video sit behind the same REST API and draw on the same plan, and the same key also covers image, speech and LLM generation.

Explore Our Other Solutions

Unlock your creative potential and scale your business with ModelsLab's comprehensive suite of AI-powered solutions.

Imagen

AI Image Generation & Tools

Generate, edit, upscale, and transform images with state-of-the-art AI models.

Explore Imagen
Audio Gen

AI Audio Generation

Text-to-speech, voice cloning, music generation, and audio processing APIs.

Explore Audio Gen
Chat

Engage Seamlessly with LLM

Access powerful language models for chatbots, content generation, and AI assistants.

Explore Chat
3D Verse

Create Stunning 3D Models

Transform images and text into 3D models with advanced AI-powered generation.

Explore 3D Verse
Plugins

Explore Plugins for Pro

Our plugins are designed to work with the most popular content creation software.

Explore PluginsLearn More
API

Build Apps with
ML
API

Use our API to build apps, generate AI art, create videos, and produce audio with ease.

API DocumentationPlayground