flux-3-video
flux-3-video is available through TTAPI. FLUX 3 Video uses the BFL official format for text-to-video, image keyframes, video continuation, and draft enhancement with per-second billing.
InputPrompttext
Models1available
FlowAsyncresult URL
Pricefrom 3.6quota / sec
Examples
Prompt to video, in context
Text-to-video · no reference images
Prompt
A handheld flashlight sweeps through a dense forest at night, catching wet leaves and pale tree trunks in the beam, grounded found-footage atmosphere, gentle forward movement, and natural low-light grain.
Prompt
Place the supplied green-cap character on a large digital billboard above a busy New York street. Preserve its bold black-and-teal line art while traffic and pedestrians move naturally below.
Prompt
Continue the supplied wildlife clip beyond its final frame: elephants move through a dry savannah at orange sunset, keeping the low documentary framing, dust, and warm horizon light consistent.
Capabilities
What you can build with flux-3-video
Text and keyframe video
Generate from text, one or more input-image keyframes, or timed keyframe pairs.
Continue existing video
Use video-to-video mode to extend the ending frames of an input MP4.
Draft enhancement
Generate a lower-cost draft, then use its draft cache to render the full-quality version.
Pricing
Per-action, in quota
Usage is metered in quota by model, action, and output settings.
ModelResolutionWorkflowTTAPI price
Flux 3 VideoHDText / image → video10.2 quota / second
Flux 3 VideoFHDText / image → video17.4 quota / second
Flux 3 VideoHDVideo → video24.6 quota / second
Flux 3 VideoFHDVideo → video31.8 quota / second
Flux 3 Video DraftHDText / image → video3.6 quota / second
Flux 3 Video DraftHDVideo → video7.2 quota / second
API request
Headers and submit parameters
Use the endpoint below with your TTAPI key and the request fields shown.
Headers
TT-API-KEYstringrequiredYour TTAPI API key.
Content-TypestringrequiredUse application/json.
Body
promptstringconditionalFree-form video prompt. Required for t2v, i2v, and v2v; draft_enhance reuses the original prompt from draft_cache.
modestringrequiredDefaults to t2v. Use t2v, i2v, v2v, or draft_enhance for the selected workflow.
keyframesstring | any[]conditionalRequired for i2v. Accepts one to ten HTTP(S) or base64 images, or ordered [second, image] pairs for timed keyframes.
start_videostringconditionalHTTP(S) URL or base64 MP4 used to continue a video in v2v mode.
draft_cachestringconditionalRequired for draft_enhance. Use the base64-encoded .bin cache from a prior draft job, or its still-valid HTTP(S) URL.
aspect_ratioenumoptionalDefaults to auto. Supported: auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, and 9:16.
durationinteger | enum<string>optionalDefaults to auto. Use an integer from 5 to 20 seconds, or auto to determine duration from the content.
resolutionenumoptionalhd or fhd; draft enhancement defaults to fhd.
versionenumoptionallatest uses the current Flux 3 version.
generate_audiobooleanoptionalGenerate synchronized audio with the video.
safety_toleranceintegeroptionalDefaults to 2. Content-safety tolerance from 0 (strictest) to 4.
draftbooleanoptionalReturn a lower-cost draft and draft_cache for a later full-quality enhancement.
webhook_urlstringoptionalTTAPI asynchronous callback URL; otherwise retrieve the completed job through the Flux result endpoint.
Integrate
From key to first result
Copy the resolved endpoint, authentication headers, and request body for this model.
curl --request POST \
--url https://api.ttapi.io/bfl/v1/flux-3-video \
--header 'TT-API-KEY: $TTAPI_KEY' \
--header 'Content-Type: application/json' \
--data '{"prompt":"A handheld flashlight sweeps through a dense forest at night, catching wet leaves and pale tree trunks in the beam.","mode":"t2v","aspect_ratio":"auto","duration":"auto","resolution":"hd","version":"latest","generate_audio":true,"safety_tolerance":2,"draft":false,"webhook_url":"https://example.com/callback"}'
