cn8/video-from-text

Turn a text prompt into a short video. Pick a model with the `model` field — Google Veo,

Learn more about Video from Text

Videoasync (job)Instant previewvideogenerationtext-to-video

Describe your video

Model

Result

Your video shows up here

Pick a model — its real durations and resolutions load automatically.

Overview

Turn a text prompt into a short video. Pick a model with the model field — Google Veo, Seedance 2.0, Kling AI, Luma Ray, Runway Gen 4 or MiniMax Hailuo — then set resolution, duration and aspect ratio, and the video is generated for you.

Call List Video Models (media-video-models) for the live catalog: every model's provider, per-second price, supported modes, resolutions and durations.

This is an asynchronous job: you submit a request, receive a job ID right away, then poll until the video is ready.

Pricing is per second of generated video and depends on the model: Veo 80 credits/s · Seedance 50 credits/s · Kling 40 credits/s · Luma 30 credits/s · Runway Gen 4.5 25 credits/s · Runway Gen 4 Turbo 10 credits/s · MiniMax Hailuo 10 credits/s (hailuo-2.3-fast 8 credits/s).

Key capabilities

Six providers, one API

The model field picks the provider: Veo (veo-3.1-generate-001 and friends), Seedance (seedance-2-0, -fast, -mini), Kling (kling-v3, kling-v2-6, kling-v2-5-turbo, kling-3-0-turbo), Luma (ray-3.2), Runway (gen4.5) or MiniMax (hailuo-2.3, -fast, hailuo-02). Same request shape for all.

Live model catalog

GET media-video-models (free) returns every model with its price, supported modes, resolutions, aspect ratios and durations — ideal for building a model picker.

Simple async flow

Submit a job, get a job ID back immediately, and poll until the video is ready.

Configurable resolution, duration & aspect ratio

Veo: 512P/1080P/4K, 4-8s, 16:9 or 9:16. Seedance/Kling: up to 1080p (Kling 4k), 3-15s, plus 1:1. Luma: 5s or 9s. Runway: 2-10s and extra ratios (4:3, 3:4, 21:9). See media-video-models for the authoritative per-model list.

Per-second, per-model pricing

Veo 80 credits/s · Seedance 50 credits/s · Kling 40 credits/s · Luma 30 credits/s · Runway Gen 4.5 25 credits/s · Gen 4 Turbo & MiniMax Hailuo 8 credits-0.10/s — billed for the generated video's duration when the job completes.

When to use it

Marketing snippets

Short promo / ad / social video from a brief text prompt.

Animated product hero

Generate a short looping hero video for landing pages.

Storyboard / pre-vis

Quick visual pre-vis from a script line before commissioning real production.

Vertical (TikTok / Reels) content

Use aspect_ratio: 9:16 to generate portrait-format videos for short-form platforms.

Input & output

input

prompt + model (required) + optional resolution, duration, aspect_ratio, generate_audio, seed

JSON body

output

A job ID right away; once complete, result_data carries the video URL, duration and metadata.

JSON (job)Video URL (in result_data)

Guides & tips

How the async flow works

1POST media-video-from-text → you get a job with an id back immediately.
2The video is generated in the background — this typically takes a few minutes.
3Poll GET /v2/jobs/{job_id} until status is completed.
4Read the finished video URL from result_data.video_url.

You're billed only when the job completes successfully. You can also configure a webhook to be notified instead of polling.

Choosing a model / provider

Call media-video-models for the live list. Snapshot:

Provider	Models	Rate	Modes	Notes
Veo (Google)	`veo-3.1-generate-001` (default), `veo-3.1-fast-generate-001`, `veo-3.1-lite-generate-001`, `veo-3.0-fast-generate-001`	80 credits/s	text, image, first-last-frame, extend	Always generates audio. 512P/1080P/4K, 4/6/8s. Only provider supporting `extend`. Legacy aliases `veo (Audio)` / `veo fast (Audio)` still accepted.
Seedance	`seedance-2-0`, `-fast`, `-mini`	50 credits/s	text, image, first-last-frame, reference-to-video	Optional audio. 720p/1080p, 4-15s, also 1:1. Only provider supporting `reference-to-video`.
Kling	`kling-v3`, `kling-v2-6`, `kling-v2-5-turbo`, `kling-3-0-turbo`	40 credits/s	text, image, first-last-frame	Up to 4k on v3/v2-6 (720p = std, 1080p = pro). `kling-3-0-turbo` has no first-last-frame.
Luma	`ray-3.2`	30 credits/s	text, image, first-last-frame	Durations 5s or 9s only.
Runway	`gen4.5`	25 credits/s	text, image	Extra aspect ratios (4:3, 3:4, 21:9), 2-10s.
Runway	`gen4_turbo`	10 credits/s	image only	Requires an input image — no text-only generation.
MiniMax	`hailuo-2.3`, `hailuo-02`	10 credits/s	text, image, first-last-frame	Cheapest text-to-video. 768P/1080P, 6s or 10s (1080P limits vary).
MiniMax	`hailuo-2.3-fast`	8 credits/s	image only	Cheapest option overall.

Rules of thumb: veo-3.1-generate-001 for final quality with audio, hailuo-2.3 for the cheapest text-to-video loop, hailuo-2.3-fast / gen4_turbo for cheap image animation, ray-3.2 / kling-v3 as mid-range options.

Aspect ratio choice

16:9 — landscape, default. Best for desktop, YouTube and TV.
9:16 — portrait. Best for TikTok, Instagram Reels and Shorts.

Specs

Latency

Asynchronous — typically 1-5 minutes (longer for 1080P or 8s clips). Jobs time out after 10 minutes.

Async

true

Rate Limit

60 req/min per API key

Max Input

Long prompts are supported

Schema

Request body

promptrequiredstring

Text description of the video to generate.

modelrequiredstring

Picks the provider and model — use a model_id / call_as / alias from media-video-models. Veo: 'veo-3.1-generate-001' (default), 'veo-3.1-fast-generate-001', 'veo-3.1-lite-generate-001', 'veo-3.0-fast-generate-001' (aliases 'veo (Audio)' / 'veo fast (Audio)' still work). Seedance: 'seedance-2-0', 'seedance-2-0-fast', 'seedance-2-0-mini'. Kling: 'kling-v3', 'kling-v2-6', 'kling-v2-5-turbo', 'kling-3-0-turbo'. Luma: 'ray-3.2'. Runway: 'gen4.5'. MiniMax: 'hailuo-2.3', 'hailuo-2.3-fast' (image-only), 'hailuo-02'.

generation_typestring · default: text-to-video

The generation process. Defaults to 'text-to-video' (this service). Other values are served by their own gateway services: 'image-to-video' (media-video-from-image), 'first-last-frame' (media-video-first-last-frame), 'extend' (media-video-extend, Veo only).

resolutionstring · default: 1080P (Veo) / 720p (others)

Output resolution — per model. Veo: '512P' / '1080P' / '4K'. Seedance & Luma: '720p' / '1080p'. Kling: '720p' / '1080p' / '4k' (720p = std mode, 1080p = pro mode). Runway: '720p' / '1080p' (Gen 4 Turbo: 720p only). See media-video-models.

durationstring · default: 4s (Veo) / 5s (others)

Video duration as a string with the 's' suffix — allowed values per model. Veo: 4/6/8s. Seedance: 4/5/6/8/10/15s. Kling: 3-15s. Luma: 5s or 9s. Runway Gen 4.5: 2-10s (Gen 4 Turbo: 5s or 10s). See media-video-models.

aspect_ratiostring · default: 16:9

Aspect ratio. Veo: '16:9' or '9:16'. Seedance/Kling/Luma also allow '1:1'. Runway additionally allows '4:3', '3:4', '21:9'.

generate_audioboolean

Seedance/Kling/Luma/Runway: generate a soundtrack with the video. Veo models always include generated audio.

seedinteger

Non-Veo providers: seed for reproducible generations (-1 = random).

Pricing

Per second of generated video, at the model's own rate. You're billed once, when the job completes successfully. media-video-models always returns the live prices.

Service	Unit	Price
Veo models (veo-3.1-, veo-3.0-)	second	80 credits/second
Seedance models (seedance-2-0*)	second	50 credits/second
Kling models (kling-*)	second	40 credits/second
Luma (ray-3.2)	second	30 credits/second
Runway Gen 4.5 (gen4.5)	second	25 credits/second
Runway Gen 4 Turbo (gen4_turbo)	second	10 credits/second
List Video Models	item	Free

Examples: a 4s Veo video = 320 credits; a 5s Kling video = 200 credits; a 5s Luma video = 150 credits; a 5s Runway Gen 4 Turbo video = 50 credits.

Within a provider, model choice affects speed/quality, not the rate — except Runway, where Gen 4 Turbo (10 credits/s) is its own rate.

If a job fails, it isn't billed — you're only charged for videos that complete successfully.

FAQ

Related models

Video from Image

Animate an existing image into a short video instead of generating from text.

Image Generation

Generate still images from text (faster, cheaper).

Dubbing

Add localized voice narration to an existing video.