Which video models can my agent use?

All 9, priced per second with a 4-second minimum and an 8-second default: seedance-lite ($0.036/s) for budget clips; runway-gen4.5 ($0.12/s), kling-video ($0.14/s, in 5-second steps), veo-3.1-fast, seedance-pro and seedance-2 ($0.15/s) in the middle; and veo-3.1 ($0.40/s) at the top. estimate_cost prices any duration before it runs, and no separate Google, ByteDance or Kuaishou accounts are needed.

How does async video generation work over MCP?

generate_video (or image_to_video) returns a job immediately; the agent polls get_generation_status until the finished video URL is ready — typically one to a few minutes. Well-behaved agents handle the loop themselves; if one gives up, ask it to check the job again.

Can I turn an existing image into video?

Yes — upload_asset pushes your still (a render, a photo, a generated image) and image_to_video animates it with a motion prompt. It's the standard path for product turntables, loops from posters, and bringing generated art to life.

What if a video job fails halfway?

You're refunded automatically — billing is per successful output only. Combined with idempotent requests and your monthly spending limit, an agent can't rack up silent failures on your credit balance.

Is per-second video actually cheap?

Compared to running these models direct, yes: no minimum spend, no GPU rental, no subscription per provider. A batch of ten 8-second clips on veo-3.1-fast is a predictable $12.00, and estimate_cost confirms it before anything runs.

Which agent setups work best for video?

claude.ai and ChatGPT (OAuth connectors) are great interactively; Claude Code and Codex shine for pipelines — e.g. generating release-notes demo videos in CI with a tightly capped API key.

Create free key

Video generationAERA

AI video generation, one tool call away.

Veo 3.1, Veo 3.1 Fast, Seedance 2, Seedance Pro & Lite, Kling 2.6, and Runway Gen-4.5 — callable from any MCP agent. Text-to-video, image-to-video, async status polling, priced per second from $0.036/s (Seedance Lite) to $0.40/s (Veo 3.1) — 4-second minimum, 8-second default.

Connect in 60 seconds

generate_video · AERA rain through NinjaChat

Veo 3.1from $1.60

Seedancefrom $0.605

Klingfrom $0.70

Paste this. Watch it run.

First ask

An 8-second aerial shot of a lighthouse in a storm, cinematic, for our title sequence.

generate_videoget_generation_status

8s video on veo-3.1 · $3.20

Use case · still → motion

Night track. Then the rain moves.

One product still. One image_to_video call. Eight seconds of weather that belongs to the shoe.

AERA · motion

Same shoe. Night track. Then the rain moves.

Models

9 video engines. Priced before they run.

Google Veo 3.1Veo 3.1

Seedance 2.0Seedance

Kling 2.6 ProKling

Google Veo 3.1 FastVeo Fast

Seedance LiteSeedance Lite

Tools

21 tools. These do the work here.

Make

generate_videoTurn a description into a video.
image_to_videoAdd motion to a still image.
get_generation_statusSee whether a video is ready to view.

Library

upload_assetUpload an image to edit or use as a reference.

Money

estimate_costCheck the price before generating.

Also on the server

create_mediagenerate_speechgenerate_imageedit_imageedit_videoextend_videosearch_creationsget_creationsave_creation_aslist_creative_entitiescompare_modelssearch_modelsprepare_scrub_videopublish_pageget_account_balanceadd_credits

Setup

Connected in about a minute.

Pick your client — the steps for the others stay one tap away.

Claude4 steps

Open connector settings
In claude.ai, go to Settings → Connectors and choose “Add custom connector”. Your account and workspace must allow custom connectors.
2
Paste the NinjaChat endpoint
Name it NinjaChat and paste the server URL — nothing else to configure.
text
```
https://www.ninjachat.ai/api/mcp
```
Sign in when prompted
Sign in to NinjaChat and approve the connection. No API key needed. Enable NinjaChat in each chat or project where you want to use it.
Ask Claude to make something
Start a new chat and enable NinjaChat in the tools menu.

Claude Code4 steps

Create an API key
Sign in at ninjachat.ai and create a key under Developers → Keys. Keys look like nj_sk_… and carry your spending limit.
2
Add the server with one command
Register NinjaChat as an HTTP MCP server. Use --scope user to make it available in every project.
bash · template
Replace nj_sk_YOUR_KEY with your API key before using this template.
```
claude mcp add --transport http --scope user ninjachat https://www.ninjachat.ai/api/mcp --header "Authorization: Bearer nj_sk_YOUR_KEY"
```
3
Verify the connection
Run /mcp inside Claude Code — ninjachat should list as connected with the ninjachat tools.
text
```
/mcp
```
Put it to work
Start a task in Claude Code, then send your prompt.