Docs

Everything to integrate images, video and audio: models, endpoints, fields, errors and webhooks. Last updated 19 October 2026.

Agent guide OpenAPI

Introduction

One REST API for ByteDance's Seedance video, Seedream images and Seed speech and sound. You pay per request from a prepaid token balance; there is no subscription.

Base URL: https://kreatorflow.ai
Auth: Authorization: Bearer <key>
Format: JSON in, JSON out: { success, data } or { success: false, code, error }
Billing: Prepaid tokens, 1 token = $0.05
Results: Async jobs: poll or receive a webhook

Everything you create is a job. Create it, keep its id, then poll GET /v1/jobs/{id} or wait for a webhook. Images take seconds, videos one to several minutes, audio returns finished in the same call.

NextQuickstart

Quickstart

Step 1
Add tokens
Top up under . You need a balance to create a key.
Step 2
Create a key
Under . It is shown once; keep it on your server.
Step 3
Make a request
Create a job, then poll it until it finishes.

1. Create an image

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: first-image-001" \
  -d '{"model":"kreator-image-pro","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"4:5"}'

2. Wait for it and download it

curl https://kreatorflow.ai/v1/jobs/$JOB_ID \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY"

# when status is "completed":
curl -L -o result.mp4 https://kreatorflow.ai/v1/jobs/$JOB_ID/output \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY"

Prefer to try first? The runs any model in the browser on your balance and gives you the exact code.

PreviousIntroduction NextAuthentication

Authentication

Send your key on every request except GET /v1/models: Authorization: Bearer kf_live_…. The x-api-key header also works.

Keys are shown once and stored hashed. Create a new one if you lose it, and revoke old ones under .
Up to 10 active keys per account. Creating a key needs a positive token balance.

Keep keys on the server

Call the API from a server or script only. Never ship a key in browser or mobile code, and never paste it into a chat with an AI assistant.

Scope	Allows
models:read	GET /v1/voices, GET /v1/storage
uploads:create	Uploads: create, complete, delete
images:create	POST /v1/images and /v1/characters
videos:create	POST /v1/videos, /v1/video-edits and /v1/lip-sync
audio:create	POST /v1/audio
jobs:read	GET /v1/jobs/{id} and its output

New keys get every scope.

PreviousQuickstart NextTokens & billing

Tokens & billing

1 token = $0.05. Every create response returns quoted_token_cost and your balance_tokens. Exact prices per model are on .

Quote. The price is worked out from model, resolution and length before anything runs.
Reserve. Those tokens are held. Without enough balance you get 402 insufficient_balance and nothing runs.
Settle. A completed job is charged finalTokenCost. A failed, canceled or blocked job is not charged; the hold is released.

Video input costs extra

A job that takes video in (reference clips or an edit's source) is billed on input plus output seconds, input at 60% of the per-second rate. The quote is then a maximum (quoted_token_cost_is_maximum: true); the real charge is finalTokenCost. Image and audio references are free.

PreviousAuthentication NextAll models

Models

Send a model alias as model. Each model has its own page with capabilities, limits, prices and real example outputs. GET /v1/models returns the live list without a key.

Seedance 2.5kreator-video-seedance-2-5from $0.143/svideo

Seedance 2.0kreator-video-pro · kreator-video-seedance-2from $0.076/svideo Seedance 2.0 Fastkreator-video-fast · kreator-video-seedance-2-fastfrom $0.058/svideo

Seedance 2.0 Minikreator-video-mini · kreator-video-seedance-2-minifrom $0.058/svideo

Wan 3.0kreator-video-wan-3from $0.07/svideo

Wan 3.0 Primekreator-video-wan-3-primefrom $0.09/svideo Kling 4.0kreator-video-kling-4from $0.18/svideo Kling 3.0kreator-video-kling-3from $0.07/svideo Kling 3.0 Omnikreator-video-kling-3-omnifrom $0.07/svideo Kling 3.0 Turbokreator-video-kling-3-turbofrom $0.09/svideo

Seedream 5.0 Prokreator-image-seedream-5-pro$0.10/imageimage

Seedream 4.5kreator-image-pro$0.10/imageimage Seedream 5.0 Flashkreator-image-seedream-5-flash$0.05/imageimage Seedream 5.0 Litekreator-image-fast$0.05/imageimage Wan 2.7 Image Prokreator-image-wan-2-7-pro$0.08/imageimage Wan 2.6kreator-image-wan-2-6$0.04/imageimage Seed Speech v3kreator-speech2 tokens / 1,000 charactersaudio Seed Audiokreator-sound2 tokens min / clipaudio Seedream 5.5kreator-image-seedream-5-5$0.06/imageimage

PreviousTokens & billing NextSeedance 2.5

Seedance 2.5

videoNewfrom $0.143/s

Newest Seedance. Clips up to 30 seconds with image, video and audio references.

kreator-video-seedance-2-5

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 9)
Reference video clips (up to 3)
Reference audio (up to 3, 15 s combined)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/videos
Resolutions: 480p, 720p, 1080p, 4K
Durations: 4s, 5s, 6s, 8s, 10s, 12s, 15s, 20s, 25s, 30s
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
References: 9 images · 3 videos · 3 audio
Prompt: up to 16,000 characters
Output: MP4 video

Pricing

Resolution	Per second	4s clip	30s clip
480p	$0.143 · 2.85 tokens	12 tokens	86 tokens
720p	$0.207 · 4.14 tokens	17 tokens	125 tokens
1080p	$0.444 · 8.88 tokens	36 tokens	267 tokens
4K	$0.999 · 19.98 tokens	80 tokens	600 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

Talking to camera, text only

720p5s16:9

Prompt

A young woman with long straight blonde hair and a white t-shirt stands outdoors in a sunny green park, talking to the camera and gesturing with her hands. Clear daylight, trees and grass behind her. Single person, head and shoulders visible.

Shown without sound.

Multi-shot 2D anime with hard cuts

720p15s16:9

Prompt (excerpt)

2D anime, vibrant, dynamic camera. A 6 shot sequence with a hard cut between every shot, no dissolves and no long single take; every shot is a different framing. Shot 1 (0-2s): Extreme close up, a thumb smashes a glowing GENERATE button on a phone, speed lines burst across the screen, his blue eyes wide. …

First 7 seconds of a 15-second clip.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-seedance-2-5-001" \
  -d '{"model":"kreator-video-seedance-2-5","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousAll models NextSeedance 2.0

Seedance 2.0

videofrom $0.076/s

Cinematic motion, reference-led shots and source-video edits, up to 4K.

kreator-video-prokreator-video-seedance-2

All aliases reach the same model at the same price.

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 9)
Reference video clips (up to 3)
Reference audio (up to 3, 15 s combined)
Saved characters (character_ids)
Video edits: transform, restyle, extend

Specs

Endpoint: POST /v1/videos
Resolutions: 480p, 720p, 1080p, 4K
Durations: 4s, 5s, 6s, 8s, 10s, 12s, 15s
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
References: 9 images · 3 videos · 3 audio
Prompt: up to 16,000 characters
Output: MP4 video

Pricing

Resolution	Per second	4s clip	15s clip
480p	$0.076 · 1.52 tokens	7 tokens	23 tokens
720p	$0.11 · 2.2 tokens	9 tokens	33 tokens
1080p	$0.24 · 4.8 tokens	20 tokens	72 tokens
4K	$0.58 · 11.6 tokens	47 tokens	174 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

Fast drone flythrough, 1080p

1080p5s16:9

Prompt

Cinematic aerial flythrough of a futuristic neon city at night, flying between towering skyscrapers covered in glowing signs, rain-slick streets reflecting blue and magenta light, dense atmosphere and light fog, smooth fast drone camera move, ultra detailed, filmic color grade

Shown without sound, scaled down for the web.

Slow push-in, atmospheric light

480p5s16:9

Prompt

A samurai in dark robes drawing a katana in a bamboo forest, leaves spiralling down around him, shafts of morning light through the mist, camera slowly pushing in

Orbiting camera move

480p5s16:9

Prompt

A knight in shining armour raising a sword to the sky on a cliff top at sunrise, red cape whipping in the wind, a castle and a river valley far below, camera slowly circling him

Tracking shot with physics

480p5s16:9

Prompt

An astronaut in a white space suit bounding across the surface of the Moon toward a lunar rover, the Earth huge in the black sky, dust kicking up with every step, camera following from the side

Short prompt, vertical

720p5s9:16

Prompt

a cat falling into a paint can full of pink paint then it jumps out covered in the paint

Shown without sound.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-pro-001" \
  -d '{"model":"kreator-video-pro","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedance 2.5 NextSeedance 2.0 Fast

Seedance 2.0 Fast

videoFastfrom $0.058/s

Quick Seedance 2.0 clips for UGC hooks, motion tests and batch variants.

kreator-video-fastkreator-video-seedance-2-fast

All aliases reach the same model at the same price.

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 9)
Reference video clips (up to 3)
Reference audio (up to 3, 15 s combined)
Saved characters (character_ids)
Video edits: transform, restyle, extend

Specs

Endpoint: POST /v1/videos
Resolutions: 480p, 720p, 1080p
Durations: 4s, 5s, 6s, 8s, 10s
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, adaptive
References: 9 images · 3 videos · 3 audio
Prompt: up to 8,000 characters
Output: MP4 video

Pricing

Resolution	Per second	4s clip	10s clip
480p	$0.058 · 1.16 tokens	5 tokens	12 tokens
720p	$0.084 · 1.69 tokens	7 tokens	17 tokens
1080p	$0.183 · 3.66 tokens	15 tokens	37 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-fast-001" \
  -d '{"model":"kreator-video-fast","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedance 2.0 NextSeedance 2.0 Mini

Seedance 2.0 Mini

videoFastfrom $0.058/s

Lowest-cost Seedance 2.0 for drafts, A/B variants and volume pipelines.

kreator-video-minikreator-video-seedance-2-mini

All aliases reach the same model at the same price.

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 9)
Reference video clips (up to 3)
Reference audio (up to 3, 15 s combined)
Saved characters (character_ids)
Video edits: transform, restyle, extend

Specs

Endpoint: POST /v1/videos
Resolutions: 480p, 720p, 1080p
Durations: 4s, 5s, 6s, 8s, 10s
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, adaptive
References: 9 images · 3 videos · 3 audio
Prompt: up to 8,000 characters
Output: MP4 video

Pricing

Resolution	Per second	4s clip	10s clip
480p	$0.058 · 1.16 tokens	5 tokens	12 tokens
720p	$0.067 · 1.35 tokens	6 tokens	14 tokens
1080p	$0.135 · 2.7 tokens	11 tokens	27 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

Four words, vertical

720p4s9:16

Prompt

a cat doing a backflip

Shown without sound.

From a character-sheet reference image

1080p5s9:16

Prompt

use the character sheet with a human tiger and make it do a cinematic fight

Made with one reference image (a character sheet, not shown). Shown without sound, scaled down for the web.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-mini-001" \
  -d '{"model":"kreator-video-mini","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedance 2.0 Fast NextWan 3.0

Wan 3.0

videoNewfrom $0.07/s

Low-cost text-to-video, frames, characters, clip and audio references, with generated sound, up to 1080p.

kreator-video-wan-3

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 10)
Reference video clips (up to 5)
Reference audio (up to 5, 15 s combined)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/videos
Resolutions: 720p, 1080p
Durations: 4s, 5s, 6s, 8s, 10s, 12s, 15s, 20s, 25s, 30s
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
References: 10 images · 5 videos · 5 audio
Prompt: up to 4,000 characters
Output: MP4 video

Pricing

Resolution	Per second	4s clip	30s clip
720p	$0.07 · 1.4 tokens	6 tokens	42 tokens
1080p	$0.14 · 2.8 tokens	12 tokens	84 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

Calm atmosphere shot

720p5s16:9

Prompt

A cosy rainy-day cafe window: raindrops race down the glass, a steaming latte with heart latte art sits on the wooden sill, blurred city lights glow outside, the camera slowly drifts closer. Warm, calm, cinematic. Sound: rain on glass, soft cafe murmur, a cup clink.

Made with generated sound; shown here without sound.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-wan-3-001" \
  -d '{"model":"kreator-video-wan-3","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedance 2.0 Mini NextWan 3.0 Prime

Wan 3.0 Prime

videoNewFastfrom $0.09/s

Wan 3.0 in fast mode: text-to-video, frames, characters, clip and audio references, with sound, up to 1080p.

kreator-video-wan-3-prime

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 10)
Reference video clips (up to 5)
Reference audio (up to 5, 15 s combined)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/videos
Resolutions: 720p, 1080p
Durations: 4s, 5s, 6s, 8s, 10s, 12s, 15s, 20s, 25s, 30s
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
References: 10 images · 5 videos · 5 audio
Prompt: up to 4,000 characters
Output: MP4 video

Pricing

Resolution	Per second	4s clip	30s clip
720p	$0.09 · 1.8 tokens	8 tokens	54 tokens
1080p	$0.19 · 3.8 tokens	16 tokens	114 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

3D animated character with a spoken line

1080p8s16:9

Prompt

3D animated film style: a small round robot with big glowing blue eyes sits on a workbench in a cosy workshop, notices a paper butterfly landing on its hand, tilts its head, and says in a cheerful little voice: "Oh! Hello there." Warm lamp light, dust in the air, soft depth of field, Pixar-like charm and detail. Sound: the robot's soft servo whirr, its voice, gentle ambient workshop hum.

Made at 1080p with generated sound; shown here at 720p without sound.

Anime action with a whip pan

1080p8s16:9

Prompt

Anime style action scene: a swordswoman with flowing silver hair in a dark kimono stands in a moonlit bamboo forest, then dashes forward and draws her katana in one blinding slash, a crescent of blue light cuts through the falling leaves. Dynamic camera whip pan, high-end anime film quality, crisp line art, glowing particles. Sound: wind through bamboo, the sharp ring of a sword being drawn, a whoosh of the slash.

Made at 1080p with generated sound; shown here at 720p without sound.

Claymation with dialogue

1080p8s16:9

Prompt

Stop-motion claymation style: a tiny clay fox in a knitted scarf sits by a crackling campfire under a starry sky, roasts a marshmallow on a stick, looks up at the camera and says in a warm little voice: "Best night ever." Handmade textures, fingerprint details in the clay, cosy warm firelight, gentle camera push in. Sound: crackling fire, crickets, the fox's voice.

Made at 1080p with generated sound; shown here at 720p without sound.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-wan-3-prime-001" \
  -d '{"model":"kreator-video-wan-3-prime","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousWan 3.0 NextKling 4.0

Kling 4.0

videoNewfrom $0.18/s

Newest Kling: clips up to 30 seconds with sound, frames, reference images and reference clips, at 1080p.

kreator-video-kling-4

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 10)
Reference video clips (up to 5)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/videos
Resolutions: 720p, 1080p
Durations: 3s, 4s, 5s, 6s, 8s, 10s, 12s, 15s, 20s, 25s, 30s
Aspect ratios: 21:9, 16:9, 9:16, 1:1
References: 10 images · 5 videos
Prompt: up to 26,000 characters
Output: MP4 video

Pricing

Resolution	Per second	3s clip	30s clip
720p	$0.18 · 3.6 tokens	11 tokens	108 tokens
1080p	$0.19 · 3.8 tokens	12 tokens	114 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-kling-4-001" \
  -d '{"model":"kreator-video-kling-4","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousWan 3.0 Prime NextKling 3.0

Kling 3.0

videoNewfrom $0.07/s

Text or first and last frame to video, a shot list, optional sound, up to 4K.

kreator-video-kling-3

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 2)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/videos
Resolutions: 720p, 1080p, 4K
Durations: 3s, 4s, 5s, 6s, 8s, 10s, 12s, 15s
Aspect ratios: 16:9, 9:16, 1:1
References: 2 images
Prompt: up to 3,072 characters
Output: MP4 video

Pricing

Resolution	Per second	3s clip	15s clip
720p	$0.10 · 2 tokens	6 tokens	30 tokens
1080p	$0.12 · 2.4 tokens	8 tokens	36 tokens
4K	$0.14 · 2.8 tokens	9 tokens	42 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-kling-3-001" \
  -d '{"model":"kreator-video-kling-3","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousKling 4.0 NextKling 3.0 Omni

Kling 3.0 Omni

videoNewfrom $0.07/s

Frames plus up to 7 reference images or characters, a reference clip or a clip to edit, a shot list, up to 4K.

kreator-video-kling-3-omni

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 7)
Reference video clips (up to 1)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/videos
Resolutions: 720p, 1080p, 4K
Durations: 3s, 4s, 5s, 6s, 8s, 10s, 12s, 15s
Aspect ratios: 16:9, 9:16, 1:1
References: 7 images · 1 videos
Prompt: up to 3,072 characters
Output: MP4 video

Pricing

Resolution	Per second	3s clip	15s clip
720p	$0.09 · 1.8 tokens	6 tokens	27 tokens
1080p	$0.10 · 2 tokens	6 tokens	30 tokens
4K	$0.12 · 2.4 tokens	8 tokens	36 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-kling-3-omni-001" \
  -d '{"model":"kreator-video-kling-3-omni","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousKling 3.0 NextKling 3.0 Turbo

Kling 3.0 Turbo

videoNewFastfrom $0.09/s

Fast Kling 3.0 clips with sound from text or a first frame, up to 1080p.

kreator-video-kling-3-turbo

Capabilities

Text to video
Image to video with a first and/or last frame
Reference images (up to 1)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/videos
Resolutions: 720p, 1080p
Durations: 3s, 4s, 5s, 6s, 8s, 10s, 12s, 15s
Aspect ratios: 16:9, 9:16, 1:1
References: 1 images
Prompt: up to 3,072 characters
Output: MP4 video

Pricing

Resolution	Per second	3s clip	15s clip
720p	$0.09 · 1.8 tokens	6 tokens	27 tokens
1080p	$0.10 · 2 tokens	6 tokens	30 tokens

Billed per whole request (rate × seconds, rounded up to a whole token). Reference clips add input seconds; see Tokens & billing.

Quick request

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-video-kling-3-turbo-001" \
  -d '{"model":"kreator-video-kling-3-turbo","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousKling 3.0 Omni NextSeedream 5.0 Pro

Seedream 5.0 Pro

imageNew$0.10/image

Most detailed stills with up to 10 references.

kreator-image-seedream-5-pro

Capabilities

Text to image
Reference images (up to 10)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/images
Aspect ratios: 1:1, 4:5, 3:4, 9:16, 16:9, 21:9
Resolution: standard
References: 10 images
Prompt: up to 8,000 characters
Output: One image per request

Pricing

$0.10/image · 2 tokens per image

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

Macro detail
16:9
Prompt
Cinematic editorial cover photograph: extreme macro of a hummingbird hovering by a flower, every feather razor sharp with iridescent detail, the left half rendered slightly softer and the right half ultra crisp, dark background, studio lighting, ultra detailed, premium magazine look. No text, no logos, no watermark.

Product shot
16:9
Prompt
Cinematic editorial cover photograph: a luxury perfume bottle floating above black water, splashes frozen mid-air around it, a single dramatic spotlight, gold and amber reflections, high-end commercial product advertising look, shallow depth of field, ultra detailed. No text, no logos, no brand names, no watermark.

Consistent character turnaround
16:9
Prompt (excerpt)
Three identical stylised 3D fox character figurines on small turntables in a dark studio, the same character shown from the front, side and three-quarter view, consistent design and colours, soft blue rim light, character turnaround sheet feel without any text.

Action still
16:9
Prompt (excerpt)
A small racing FPV drone diving low through a narrow red-rock canyon at blue hour, strong motion blur on the canyon walls, glowing blue propeller trails, dynamic low angle, cinematic action still.

Light and materials
16:9
Prompt
Cinematic editorial cover photograph: a glowing glass prism on a black surface splitting a beam of white light into a vivid spectrum that paints a miniature surreal landscape on the table, studio lighting, ultra detailed, premium magazine look, elegant and minimal. No text, no logos, no watermark.

2D illustration
16:9
Prompt
Bright colourful 2D cartoon illustration for an editorial cover: a cheerful cartoon fox and a small round robot sitting on a hill under a giant play-button shaped sun, rolling green hills and fluffy clouds, bold clean outlines, rich saturated colours, polished animation studio quality, wide 16:9 composition. No text, no logos, no watermark.

Quick request

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-image-seedream-5-pro-001" \
  -d '{"model":"kreator-image-seedream-5-pro","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"16:9"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousKling 3.0 Turbo NextSeedream 4.5

Seedream 4.5

image$0.10/image

Product shots, character references and production stills, up to 14 references.

kreator-image-pro

Capabilities

Text to image
Reference images (up to 14)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/images
Aspect ratios: 1:1, 4:5, 3:4, 9:16, 16:9, 21:9
Resolution: high
References: 14 images
Prompt: up to 8,000 characters
Output: One image per request

Pricing

$0.10/image · 2 tokens per image

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

Cinematic scene
standard16:9
Prompt
A hyperrealistic bioluminescent jellyfish drifting in the deep dark ocean, glowing cyan and magenta tendrils, volumetric god rays, tiny glowing particles, cinematic, ultra detailed

Short prompt, photo look
standard1:1
Prompt
A golden retriever puppy sitting in a sunny meadow, soft morning light, shallow depth of field

3D mascot
standard1:1
Prompt
a stylised 3D cartoon character of a cheerful robot mascot, full body, standing, plain background

Character on a plain background
standard1:1
Prompt
a stylised 3D cartoon character of a friendly astronaut, full body, standing pose, plain studio background

Quick request

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-image-pro-001" \
  -d '{"model":"kreator-image-pro","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"16:9"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedream 5.0 Pro NextSeedream 5.0 Flash

Seedream 5.0 Flash

imageNewFast$0.05/image

Low-cost, low-latency drafts and high-volume variants.

kreator-image-seedream-5-flash

Capabilities

Text to image
Reference images (up to 10)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/images
Aspect ratios: 1:1, 4:5, 3:4, 9:16, 16:9
Resolution: standard
References: 10 images
Prompt: up to 4,000 characters
Output: One image per request

Pricing

$0.05/image · 1 token per image

Quick request

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-image-seedream-5-flash-001" \
  -d '{"model":"kreator-image-seedream-5-flash","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"16:9"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedream 4.5 NextSeedream 5.0 Lite

Seedream 5.0 Lite

imageFast$0.05/image

Fast drafts, social visuals and template previews.

kreator-image-fast

Capabilities

Text to image
Reference images (up to 4)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/images
Aspect ratios: 1:1, 4:5, 3:4, 9:16, 16:9
Resolution: standard
References: 4 images
Prompt: up to 4,000 characters
Output: One image per request

Pricing

$0.05/image · 1 token per image

Quick request

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-image-fast-001" \
  -d '{"model":"kreator-image-fast","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"16:9"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedream 5.0 Flash NextWan 2.7 Image Pro

Wan 2.7 Image Pro

imageNew$0.08/image

Generate or edit images by instruction with up to 9 references, up to 4K.

kreator-image-wan-2-7-pro

Capabilities

Text to image
Reference images (up to 9)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/images
Aspect ratios: 1:1, 4:3, 3:4, 3:2, 2:3, 9:16, 16:9, 21:9
Resolution: 2k, 1k, 4k
References: 9 images
Prompt: up to 5,000 characters
Output: One image per request

Pricing

$0.08/image · 1.6 tokens per image

Quick request

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-image-wan-2-7-pro-001" \
  -d '{"model":"kreator-image-wan-2-7-pro","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"16:9"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeedream 5.0 Lite NextWan 2.6

Wan 2.6

imageNewFast$0.04/image

Low-cost text to image at about 1.3K.

kreator-image-wan-2-6

Capabilities

Text to image
Reference images (up to 0)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/images
Aspect ratios: 1:1, 4:3, 3:4, 3:2, 2:3, 9:16, 16:9, 21:9
Resolution: standard
References: —
Prompt: up to 2,000 characters
Output: One image per request

Pricing

$0.04/image · 0.8 tokens per image

Quick request

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-image-wan-2-6-001" \
  -d '{"model":"kreator-image-wan-2-6","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"16:9"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousWan 2.7 Image Pro NextSeed Speech v3

Seed Speech v3

audioFast2 tokens / 1,000 characters

Text to speech with preset and cloned voices.

kreator-speech

Capabilities

Text to speech
Voices from GET /v1/voices
Finished in the same call

Specs

Endpoint: POST /v1/audio
Input: text + voice
Prompt: up to 5,000 characters
Output: Audio file, returned finished

Pricing

2 tokens / 1,000 characters · 2 tokens / 1,000 characters

Quick request

curl -X POST https://kreatorflow.ai/v1/audio \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-speech-001" \
  -d '{"model":"kreator-speech","voice":"<voice id from GET /v1/voices>","text":"Welcome to KreatorFlow."}'

Note

The response already contains the finished job with providerOutputUrl. No polling needed.

PreviousWan 2.6 NextSeed Audio

Seed Audio

audioFast2 tokens min / clip

Sound effects, ambient beds and music from a prompt.

kreator-sound

Capabilities

Sound effects and music from a prompt
Finished in the same call

Specs

Endpoint: POST /v1/audio
Input: prompt
Prompt: up to 3,000 characters
Output: Audio file, returned finished

Pricing

2 tokens min / clip · 2 tokens / clip minimum, then by output length

Examples

Real outputs from this model, made on KreatorFlow. Each shows the prompt and the request that sets up the same job.

Narrated line

Prompt

A cheerful female narrator says: Welcome to KreatorFlow audio.

Quick request

curl -X POST https://kreatorflow.ai/v1/audio \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-sound-001" \
  -d '{"model":"kreator-sound","prompt":"Soft rain on a tin roof with distant thunder"}'

Note

The response already contains the finished job with providerOutputUrl. No polling needed.

PreviousSeed Speech v3 NextSeedream 5.5

Seedream 5.5

imageNew$0.06/image

Precise multi-round edits, realistic multi-image blends, small and multilingual text

kreator-image-seedream-5-5

Capabilities

Text to image
Reference images (up to 10)
Saved characters (character_ids)

Specs

Endpoint: POST /v1/images
Aspect ratios: 1:1, 4:5, 3:4, 9:16, 16:9, 21:9
Resolution: standard
References: 10 images
Prompt: up to 8,000 characters
Output: One image per request

Pricing

$0.06/image · 1.2 tokens per image (was $0.10, 40% off until 11 December)

Quick request

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: kreator-image-seedream-5-5-001" \
  -d '{"model":"kreator-image-seedream-5-5","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"16:9"}'

Note

The response is a job. Poll GET /v1/jobs/{id} every few seconds, or send webhook_url.

PreviousSeed Audio NextImages

Images

POST/v1/images

curl -X POST https://kreatorflow.ai/v1/images \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: image-001" \
  -d '{"model":"kreator-image-pro","prompt":"studio product photo of red boots on wet asphalt","aspect_ratio":"4:5"}'

Errors

Code	HTTP	What to do
invalid_api_key	401	Send Authorization: Bearer <key>. Create a new key if it was revoked.
insufficient_balance	402	Top up under Billing. data.top_up_url links there.
prompt_required	400	Add a prompt.
prompt_too_long	400	Shorten it; GET /v1/models lists maxPromptChars.
model_not_found	400	Use an id from GET /v1/models.
aspect_ratio_not_supported	400	Send one of the model's aspect ratios.
reference_upload_not_found	404	Re-upload the file and use the new upload_id.
character_pending	409	Wait retry_after seconds, then retry.
rate_limited	429	Wait the seconds in the Retry-After header.

References can be public HTTPS links or upload_ids from Uploads. For a face you reuse, register a character.

PreviousSeedream 5.5 NextVideos

Videos

POST/v1/videos

curl -X POST https://kreatorflow.ai/v1/videos \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: video-001" \
  -d '{"model":"kreator-video-fast","prompt":"a paper boat drifting across a calm pond at sunrise","duration":5,"aspect_ratio":"16:9","resolution":"720p"}'

Errors

Code	HTTP	What to do
invalid_api_key	401	Send Authorization: Bearer <key>. Create a new key if it was revoked.
insufficient_balance	402	Top up under Billing. data.top_up_url links there.
prompt_required	400	Add a prompt.
model_not_found	400	Use an id from GET /v1/models.
duration_not_supported	400	Send one of the model's durations.
aspect_ratio_not_supported	400	Send one of the model's aspect ratios.
resolution_not_supported	400	Send one of the model's resolutions.
reference_combination_not_supported	400	Send one kind per request.
character_pending	409	Wait retry_after seconds, then retry.
rate_limited	429	Wait the seconds in the Retry-After header.

Send only durations, aspect ratios and resolutions the model lists; anything else returns a 400 instead of a silent default.
Start/end frames (first_frame_*, last_frame_*) can't be combined with reference images or character_ids in one request.
Audio references may total 15 seconds across all of them. Keep video references short: input seconds are billed (see Tokens & billing).

Reference clips: fallback or strict

With video_reference_mode: "fallback" (default) a clip reference may be dropped to get you an output; the job then has referenceDropped: true. Send "strict" when the clip must be used.

PreviousImages NextVideo edits

Video edits

Transform, restyle or extend an existing clip with Seedance 2. The source is used strictly and framing defaults to adaptive.

POST/v1/video-edits

curl -X POST https://kreatorflow.ai/v1/video-edits \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: edit-001" \
  -d '{"model":"kreator-video-seedance-2","edit_mode":"restyle","source_video_upload_id":"tos:kreatorflow/uploads/…/clip.mp4","prompt":"turn the scene into a rainy neon night","resolution":"720p"}'

The same edit also works on POST /v1/videos with intent: "video_edit". Upload the source with upload_type: "VIDEO"; MP4/MOV (H.264/AAC) under 50 MB works best.

Errors

Code	HTTP	What to do
insufficient_balance	402	Top up under Billing. data.top_up_url links there.
video_edit_model_not_supported	400	Use kreator-video-seedance-2, kreator-video-fast or kreator-video-mini.
video_edit_source_required	400	Send source_video_upload_id or video.
video_edit_source_not_found	404	Upload the clip with upload_type VIDEO.
resolution_not_supported	400	Send one of the model's resolutions.
rate_limited	429	Wait the seconds in the Retry-After header.

PreviousVideos NextAudio

Audio

Text to speech (kreator-speech) and sound or music from a prompt (kreator-sound). The call returns the job already finished, with providerOutputUrl.

POST/v1/audio

curl -X POST https://kreatorflow.ai/v1/audio \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: speech-001" \
  -d '{"model":"kreator-speech","voice":"<voice id>","text":"Welcome to KreatorFlow."}'

GET/v1/voices

curl https://kreatorflow.ai/v1/voices \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY"

Errors

Code	HTTP	What to do
invalid_api_key	401	Send Authorization: Bearer <key>. Create a new key if it was revoked.
insufficient_balance	402	Top up under Billing. data.top_up_url links there.
prompt_required	400	Add a prompt.
voice_required	400	Use an id from GET /v1/voices.
rate_limited	429	Wait the seconds in the Retry-After header.

PreviousVideo edits NextLip Sync

Lip Sync

One photo and one voice track become a video of that person speaking it (kreator-lipsync, OmniHuman 2.0). The video is as long as the audio, 4 seconds to 5 minutes. Upload the photo (upload_type: "IMAGE") and the audio (upload_type: "AUDIO", MP3 or WAV) first, or pass a finished /v1/audio job as audio_job_id.

POST/v1/lip-sync

curl -X POST https://kreatorflow.ai/v1/lip-sync \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: lipsync-001" \
  -d '{"image_upload_id":"tos:kreatorflow/uploads/…/presenter.png","audio_upload_id":"tos:kreatorflow/uploads/…/voiceover.mp3","prompt":"smiles and nods while talking, steady camera","resolution":"720p"}'

Consent and paid accounts

Only use photos of yourself or of people who agreed, as the Terms and Acceptable Use Policy require. Lip Sync needs a paid account (a card top-up). The photo is screened and the audio is measured before anything is reserved.

4 tokens per started second of audio at 720p, 6 at 1080p. The response returns billed_seconds. Lip Sync renders one job at a time: while it is busy your job stays queued with its tokens held and starts when a slot frees, so keep polling the job.

Errors

Code	HTTP	What to do
paid_account_required	403	Top up with a card under Billing.
paid_check_unavailable	503	Send the same request again in a moment.
insufficient_balance	402	Top up under Billing. data.top_up_url links there.
image_upload_not_found	404	Upload the photo with upload_type IMAGE and use the new upload_id.
audio_upload_not_found	404	Upload the track with upload_type AUDIO, or use a completed audio job id.
audio_too_short	400	Send 4 seconds to 5 minutes of audio.
audio_too_long	400	Trim it to 5 minutes or less.
audio_unreadable	400	Use an MP3 or WAV under 15 MB.
image_rejected	422	Use a different photo.
rate_limited	429	Wait the seconds in the Retry-After header.

PreviousAudio NextCharacters

Characters

Register a face or character once and reuse it by id. It is reviewed once (about 20 seconds for an image); after that, jobs use it directly with no per-job review.

Step 1
Upload the image
Use the upload flow.
Step 2
Register it
POST /v1/characters with mode: "register" and the upload_id.
Step 3
Wait, then use it
Poll until status is ready, then pass character_ids.

POST/v1/characters

curl -X POST https://kreatorflow.ai/v1/characters \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"mode":"register","name":"Mira","upload_id":"tos:kreatorflow/uploads/…/mira.png"}'

Real people in references

Photos and clips of real people you have the rights and consent to use work as references. A new reference is reviewed once on its first use (about 20 seconds for an image, up to about 2.5 minutes for a video clip); reuse the same upload_id and later jobs skip the review. For a face you use often, register it here as a character. Celebrity, public-figure and impersonation requests, and prompts that ask to match someone's identity, are refused.

GET/v1/characters

GET/v1/characters/{character_id}

DELETE/v1/characters/{character_id}

When	You get	Do this
`status: pending`	Still in review	Poll GET /v1/characters/{id}
`409 character_pending`	Used before ready; nothing reserved	Wait retry_after seconds, retry
`422 character_registration_failed`	The image didn't pass review	Register again with another image

PreviousLip Sync NextUploads

Uploads

Private reference files go through a three-step upload. Public HTTPS links can skip this and go straight into the reference fields.

# 1. Reserve an upload
curl -X POST https://kreatorflow.ai/v1/uploads \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"file_name":"ref.png","mime_type":"image/png","file_size":482133,"upload_type":"IMAGE"}'

# 2. PUT the bytes to direct_upload_url with exactly direct_upload_headers
curl -X PUT "$DIRECT_UPLOAD_URL" -H "Content-Type: image/png" --data-binary @ref.png

# 3. Finish it
curl -X POST https://kreatorflow.ai/v1/uploads/complete \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"upload_id":"'"$UPLOAD_ID"'"}'

POST/v1/uploads

Response

{
  "success": true,
  "data": {
    "upload_id": "tos:kreatorflow/uploads/workspace-id/1700000000000-reference.png",
    "direct_upload_url": "https://…signed-storage-url…",
    "direct_upload_headers": {
      "Content-Type": "image/png"
    },
    "direct_upload_expires_at": "2026-10-17T12:10:00.000Z",
    "complete_url": "https://kreatorflow.ai/v1/uploads/complete",
    "upload_method": "PUT",
    "upload_mode": "direct"
  }
}

POST/v1/uploads/complete

DELETE/v1/uploads/{upload_id}

Errors

Code	HTTP	What to do
upload_invalid	400	Send file_name, mime_type, file_size and upload_type.
upload_too_large	413	Images 50 MB, video 200 MB, audio 15 MB.
upload_larger_than_declared	413	Create a new upload with the real file_size and PUT the file again.
storage_quota_exceeded	402	Delete old uploads or add a Storage Pack.
upload_not_found	404	Create the upload again.
upload_in_use	409	Delete it after the job finishes.

Type	upload_type	Max size
Image	IMAGE	50 MB
Video	VIDEO	200 MB
Audio	AUDIO	15 MB

Send exactly direct_upload_headers on the PUT; the signed URL lasts about 5 minutes (10 with signed_size: true). A new reference is reviewed on its first use (about 20 s for an image, up to about 2.5 min for a clip); reuse the same upload_id and later jobs skip it.

PreviousCharacters NextJobs & polling

Jobs & polling

Status	Meaning
created · queued · processing	Still running. Poll every 3 seconds.
completed	Done. Download providerOutputUrl with your key.
failed	Read errorCode and errorMessage. Not charged.
canceled · blocked	Stopped (blocked = e.g. no balance). Not charged.

GET/v1/jobs/{job_id}

curl https://kreatorflow.ai/v1/jobs/3f0c0e0a-0000-4000-8000-000000000000 \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY"

The job object

Response fields

idstring

Job id. Poll GET /v1/jobs/{id} with it.

mediaTypestring

image, video or audio.

modelIdstring

The model alias you sent.

statusstring

created, queued, processing, completed, failed, canceled or blocked.

quotedTokenCostnumber

Tokens quoted (and reserved) when the job was created.

finalTokenCostnumber

Tokens charged. 0 until the job completes, and for any job that does not complete.

providerOutputUrlstring | null

When completed: the output link. Download it with your key.

providerOutputExpiresAtstring | null

When the output link stops working. Keep your own copy.

errorCodestring | null

Why the job failed (see Errors).

errorMessagestring | null

A readable explanation of the failure.

errorDataobject

Extra detail for some errors, e.g. retry_after for reference_asset_pending.

referenceDroppedboolean

Video, fallback mode only: true when a reference clip could not be used.

warningsarray

Non-fatal notes about the request, when there are any.

createdAt · updatedAt · completedAtstring

ISO timestamps.

GET/HEAD/v1/jobs/{job_id}/output

Output links need the same Authorization header and expire (providerOutputExpiresAt). Save a copy to your own storage.

PreviousUploads NextWebhooks

Webhooks

Add webhook_url (HTTPS) to any create request. When the job ends we POST the event below. Your signing secret is under . Keep polling as a fallback.

Events: job.completed · job.failed · job.canceled · job.blocked
Delivery: 3 attempts, 4 s timeout each
Headers: X-KreatorFlow-Event-Id, -Event-Type, -Job-Id, -Timestamp, -Signature
Signature: t=<unix>,v1=hex(HMAC-SHA256(secret, `${t}.${rawBody}`))

Event body

{
  "object": "kreatorflow.developer_webhook_event",
  "id": "evt_<job_id>_completed",
  "type": "job.completed",
  "created": "2026-10-17T12:01:10.000Z",
  "data": {
    "job": {
      "id": "<job_id>",
      "status": "completed",
      "providerOutputUrl": "https://kreatorflow.ai/v1/jobs/<job_id>/output"
    }
  }
}

Verify the signature

import { createHmac, timingSafeEqual } from "node:crypto";

// rawBody: the request body exactly as received (a string, before JSON.parse)
export function verifyKreatorFlowWebhook(rawBody, signatureHeader, secret) {
  const parts = Object.fromEntries(signatureHeader.split(",").map((part) => part.split("=")));
  const expected = createHmac("sha256", secret).update(`${parts.t}.${rawBody}`).digest("hex");
  const fresh = Math.abs(Date.now() / 1000 - Number(parts.t)) < 300; // reject replays older than 5 min
  return fresh && expected.length === parts.v1?.length && timingSafeEqual(Buffer.from(expected), Buffer.from(parts.v1));
}

PreviousJobs & polling NextStorage

Storage

Your uploads count toward a storage allowance: 5 GB free, more with a Storage Pack (see ). Generated outputs are stored but never count. Delete an upload to free its space.

GET/v1/storage

curl https://kreatorflow.ai/v1/storage \
  -H "Authorization: Bearer $KREATORFLOW_API_KEY"

Field	Meaning
quota_used_bytes	Your uploads: what counts against the allowance
included_bytes	Your allowance (pack or free)
outputs_bytes	Generated outputs (shown, not counted)
used_bytes	Everything stored
has_pack · tier	Whether a Storage Pack is active, and which

Over the allowance, new uploads return 402 storage_quota_exceeded.

PreviousWebhooks NextErrors

Errors

Every error has a stable code; branch on that, not the message. Some include extra fields in data (e.g. retry_after, top_up_url).

402 response

{
  "success": false,
  "code": "insufficient_balance",
  "error": "Not enough tokens: this job needs 21 and 4 are available. Add developer tokens before submitting generation jobs."
}

Request errors

Code	HTTP	Meaning	What to do
invalid_api_key	401	The key is missing, wrong or revoked.	Send Authorization: Bearer <key>. Create a new key if it was revoked.
insufficient_balance	402	Not enough tokens for the quoted cost.	Top up under Billing. data.top_up_url links there.
storage_quota_exceeded	402	The upload would go over your storage allowance.	Delete old uploads or add a Storage Pack.
account_not_active	403	The developer account is suspended.	Contact support.
scope_missing	403	The key lacks the scope this endpoint needs.	Use a key with that scope.
job_not_found	404	No job with that id for this key's account.	Check the id; jobs are private to the account.
character_not_found	404	The character id or slug doesn't exist.	List characters with GET /v1/characters.
upload_not_found	404	The upload_id doesn't exist in this workspace.	Create the upload again.
reference_upload_not_developer	400	The upload_id is a file uploaded on the website, not through the API.	Upload this file with /v1/uploads first.
reference_image_wrong_type	400	A reference image or frame is a video, audio or document file (data.field names it).	Send JPEG, PNG or WebP; send clips as reference_video_upload_ids.
reference_upload_not_found	404	A reference upload_id doesn't exist in this workspace.	Re-upload the file and use the new upload_id.
video_edit_source_not_found	404	The source_video_upload_id doesn't exist.	Upload the clip with upload_type VIDEO.
upload_in_use	409	A queued or processing job still uses this upload.	Delete it after the job finishes.
character_pending	409	The character is still being reviewed. Nothing was reserved.	Wait retry_after seconds, then retry.
upload_too_large	413	The file is over the size limit for its type.	Images 50 MB, video 200 MB, audio 15 MB.
upload_larger_than_declared	413	The stored file is more than 10% larger than the file_size you declared. It was deleted.	Create a new upload with the real file_size and PUT the file again.
character_registration_failed	422	The character image did not pass review.	Register the character again with another image.
rate_limited	429	Too many requests in the window.	Wait the seconds in the Retry-After header.
prompt_required	400	No prompt (or text for speech) was sent.	Add a prompt.
prompt_too_long	400	The prompt is over the model's limit.	Shorten it; GET /v1/models lists maxPromptChars.
model_not_found	400	Unknown model alias.	Use an id from GET /v1/models.
duration_not_supported	400	The model doesn't offer that length.	Send one of the model's durations.
aspect_ratio_not_supported	400	The model doesn't offer that aspect ratio.	Send one of the model's aspect ratios.
resolution_not_supported	400	The model doesn't offer that resolution.	Send one of the model's resolutions.
reference_combination_not_supported	400	These references can't be combined, e.g. start/end frames with reference images.	Send one kind per request.
upload_invalid	400	The upload request is missing a field or has a bad type.	Send file_name, mime_type, file_size and upload_type.
video_edit_model_not_supported	400	Video edits need a Seedance 2.0 family model (2.0, Fast or Mini).	Use kreator-video-seedance-2, kreator-video-fast or kreator-video-mini.
video_edit_source_required	400	A video edit needs a source clip.	Send source_video_upload_id or video.
voice_required	400	Speech needs a voice.	Use an id from GET /v1/voices.
paid_account_required	403	Lip Sync needs a paid account. Nothing was charged.	Top up with a card under Billing.
paid_check_unavailable	503	The account's payment status could not be read just now. Nothing was charged.	Send the same request again in a moment.
audio_reference_too_long	400	The audio references add up to more than the model accepts in total (15 s on Seedance 2.0, 30 s on 2.5). Nothing was charged.	Trim or remove audio so all audio references combined fit.
image_upload_not_found	404	The image_upload_id isn't in this workspace.	Upload the photo with upload_type IMAGE and use the new upload_id.
audio_upload_not_found	404	The audio_upload_id isn't in this workspace (or audio_job_id isn't a completed /v1/audio job).	Upload the track with upload_type AUDIO, or use a completed audio job id.
audio_too_short	400	The voice track is under 4 seconds. Nothing was charged.	Send 4 seconds to 5 minutes of audio.
audio_too_long	400	The voice track is over 5 minutes. Nothing was charged.	Trim it to 5 minutes or less.
audio_unreadable	400	The audio couldn't be read. Nothing was charged.	Use an MP3 or WAV under 15 MB.
image_rejected	422	The photo didn't pass the safety check. Nothing was charged.	Use a different photo.
internal_error	500	Something failed on our side.	Retry with the same Idempotency-Key.
dispatch_disabled	503	Generation is paused for this environment.	Retry later.

Job errors

The job ends failed with this errorCode. Reserved tokens are released.

errorCode	Meaning	What to do
reference_rejected	A reference did not pass screening.	For a real person you have consent to use, register the face as a character (reviewed once, then reused), or use another image.
reference_asset_pending	A new reference is still being reviewed.	Wait errorData.retry_after seconds, resubmit with the same upload_id and a new Idempotency-Key.
reference_asset_failed	The reference did not pass review.	Use a different file.
generation_rejected	The request was declined (prompt, reference or options).	Change the prompt, reference or options, then retry.
generation_timeout	The job did not finish in time; reserved tokens are released.	Retry with a new Idempotency-Key.
generation_unavailable	The model is temporarily unavailable.	Retry later or pick another model.
generation_failed	The generation failed.	Retry; change the prompt or references if it repeats.

PreviousStorage NextRate limits & retries

Rate limits & retries

Over a limit you get 429 rate_limited with a Retry-After header in seconds.

When a model is busy, a new job waits in line (queued, tokens reserved) and starts when a slot frees. One account can have up to 5 Seedance 2.5 jobs and up to 2 Lip Sync jobs waiting at once. Past that the create request also answers 429 rate_limited and nothing is reserved. Jobs already rendering do not count.

Idempotency

Send a unique Idempotency-Key header with every create. If a request times out, retry with the same key: you get the original job back (idempotent_replay: true) instead of paying twice. Use a new key for a new job, including after a failure.

PreviousErrors NextMCP server

MCP server

Let an AI client call the API as tools. Tools run through the same jobs and token balance as HTTP.

Local files: create_upload returns an upload_id and a ready curl command; run it, call confirm_upload, then pass the upload_id as reference_image_upload_ids, first_frame_upload_id, reference_video_upload_ids or image_upload_id. Same limits as POST /v1/uploads; uploading is free.

Endpoint: https://kreatorflow.ai/v1/mcp
Transport: Streamable HTTP
Auth: Authorization: Bearer <key>

list_modelscreate_uploadconfirm_uploadgenerate_imagegenerate_videoedit_videolip_synclist_voicesgenerate_speechgenerate_soundget_jobget_balance

# Claude Code
claude mcp add --transport http kreatorflow https://kreatorflow.ai/v1/mcp \
  --header "Authorization: Bearer $KREATORFLOW_API_KEY"

Using KreatorFlow with your plan credits instead of API tokens? Connect https://kreatorflow.ai/api/mcp with OAuth; setup guide.

PreviousRate limits & retries NextAI agents

AI agents

Build with ChatGPT, Claude, Cursor or Codex

One guide your assistant can follow. Save your key as KREATORFLOW_API_KEY where the code runs (never paste it into a chat), give the assistant the guide, then say what to build.

Download .md

Custom GPT or agent framework? Import the OpenAPI link and choose API key auth, type Bearer.

docs.json OpenAPI Read the guide

PreviousMCP server