> ## Documentation Index > Fetch the complete documentation index at: https://docs.bfl.ml/llms.txt > Use this file to discover all available pages before exploring further. # Generate a video with FLUX 3 > Submits a video generation task to FLUX 3 via the harness. The mode is explicit: t2v (`text-to-video`), i2v (`image-continuation`, keyframes), v2v (`video-continuation`, start_video), or draft_enhance (`draft-enhance`, full-quality render of a prior draft's `draft_cache`); the spelled-out aliases are accepted anywhere the short key is. export const ModeDeepLinks = ({modes}) => { useEffect(() => { if (!Array.isArray(modes) || !modes.length) return undefined; const press = el => { const opts = { bubbles: true, cancelable: true }; el.dispatchEvent(new PointerEvent("pointerdown", { ...opts, pointerId: 1 })); el.dispatchEvent(new MouseEvent("mousedown", opts)); el.dispatchEvent(new PointerEvent("pointerup", { ...opts, pointerId: 1 })); el.dispatchEvent(new MouseEvent("mouseup", opts)); el.dispatchEvent(new MouseEvent("click", opts)); }; const modeTabs = () => { const wanted = new Set(modes.map(m => m.tab)); const lists = Array.from(document.querySelectorAll('[role="tablist"]')).filter(list => Array.from(list.querySelectorAll('[role="tab"]')).filter(b => wanted.has(b.textContent.trim())).length >= 2); return lists.flatMap(list => Array.from(list.querySelectorAll('[role="tab"]'))); }; const exampleTrigger = () => { const names = new Set(modes.map(m => m.example)); return Array.from(document.querySelectorAll("button")).find(b => b.getAttribute("aria-haspopup") && names.has(b.textContent.trim())); }; const HIDE_MENU_STYLE_ID = "mode-deep-links-hide-menu"; const hideMenus = () => { if (document.getElementById(HIDE_MENU_STYLE_ID)) return; const style = document.createElement("style"); style.id = HIDE_MENU_STYLE_ID; style.textContent = '[data-radix-popper-content-wrapper], [role="menu"], [role="listbox"] { visibility: hidden !important; }'; document.head.appendChild(style); }; const unhideMenus = () => { document.getElementById(HIDE_MENU_STYLE_ID)?.remove(); }; const selectExample = (name, attempt = 0) => { const trigger = exampleTrigger(); if (!trigger) { if (attempt < 6) window.setTimeout(() => selectExample(name, attempt + 1), 300); return; } if (trigger.textContent.trim() === name) return; hideMenus(); press(trigger); window.setTimeout(() => { const items = Array.from(document.querySelectorAll('[role="menuitem"], [role="option"], [role="menuitemradio"]')); const item = items.find(i => i.textContent.trim() === name); if (item) { press(item); window.setTimeout(() => { unhideMenus(); const now = exampleTrigger(); if (now && now.textContent.trim() !== name && attempt < 6) { selectExample(name, attempt + 1); } }, 250); } else { document.dispatchEvent(new KeyboardEvent("keydown", { key: "Escape", bubbles: true })); unhideMenus(); if (attempt < 6) { window.setTimeout(() => selectExample(name, attempt + 1), 300); } } }, 150); }; const selectMode = (mode, {writeHash = true, switchTab = true} = {}) => { if (switchTab) { const tab = modeTabs().find(b => b.textContent.trim() === mode.tab); if (tab && tab.getAttribute("aria-selected") !== "true") press(tab); } if (writeHash) { window.history.replaceState(null, "", "#" + mode.slug); } selectExample(mode.example); }; const onClick = event => { const tab = event.target.closest('[role="tab"]'); if (!tab) return; const mode = modes.find(m => m.tab === tab.textContent.trim()); if (mode) selectMode(mode, { switchTab: false }); }; document.addEventListener("click", onClick, true); const applyHash = () => { const slug = window.location.hash.replace(/^#/, ""); const mode = modes.find(m => m.slug === slug); if (mode) selectMode(mode, { writeHash: false }); }; window.addEventListener("hashchange", applyHash); const settle = window.setTimeout(applyHash, 600); return () => { document.removeEventListener("click", onClick, true); window.removeEventListener("hashchange", applyHash); window.clearTimeout(settle); }; }, []); return null; }; ## OpenAPI ````yaml https://api.bfl.ai/openapi.json POST /v1/flux-3-video openapi: 3.1.0 info: title: BFL API description: Authorize with an API key from your user profile. version: 0.0.1 servers: - url: https://api.bfl.ai description: BFL API security: [] tags: - name: Models description: >- Generation task endpoints. These endpoints allow you to submit generation tasks. - name: Utility description: >- These utility endpoints allow you to check the results of submitted tasks and to manage your finetunes. paths: /v1/flux-3-video: post: tags: - Models summary: Generate a video with FLUX 3. description: >- Submits a video generation task to FLUX 3 via the harness. The mode is explicit: t2v (`text-to-video`), i2v (`image-continuation`, keyframes), v2v (`video-continuation`, start_video), or draft_enhance (`draft-enhance`, full-quality render of a prior draft's `draft_cache`); the spelled-out aliases are accepted anywhere the short key is. operationId: flux_3_video_v1_flux_3_video_post requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/Flux3VideoInputsBody' responses: '200': description: Successful Response content: application/json: schema: anyOf: - $ref: '#/components/schemas/AsyncResponse' - $ref: '#/components/schemas/AsyncWebhookResponse' title: Response Flux 3 Video V1 Flux 3 Video Post '422': description: Validation Error content: application/json: schema: $ref: '#/components/schemas/HTTPValidationError' security: - APIKeyHeader: [] components: schemas: Flux3VideoInputsBody: oneOf: - $ref: '#/components/schemas/Flux3VideoT2VInputs' - $ref: '#/components/schemas/Flux3VideoI2VInputs' - $ref: '#/components/schemas/Flux3VideoV2VInputs' - $ref: '#/components/schemas/Flux3VideoDraftEnhanceInputs' title: Flux3VideoInputsBody description: |- Route body wrapper: FastAPI drops the union's discriminator when the bare union is the top-level body param, so 422s enumerate every mode class instead of the one the client sent. The RootModel keeps discrimination inside pydantic; routes unwrap `.root`. discriminator: propertyName: mode mapping: draft_enhance: $ref: '#/components/schemas/Flux3VideoDraftEnhanceInputs' i2v: $ref: '#/components/schemas/Flux3VideoI2VInputs' t2v: $ref: '#/components/schemas/Flux3VideoT2VInputs' v2v: $ref: '#/components/schemas/Flux3VideoV2VInputs' AsyncResponse: properties: id: type: string title: Id polling_url: type: string title: Polling Url cost: anyOf: - type: number - type: 'null' title: Cost description: Cost in credits for this request input_mp: anyOf: - type: number - type: 'null' title: Input Mp description: Input megapixels (2 decimal places) output_mp: anyOf: - type: number - type: 'null' title: Output Mp description: Output megapixels (2 decimal places) type: object required: - id - polling_url title: AsyncResponse AsyncWebhookResponse: properties: id: type: string title: Id status: type: string title: Status webhook_url: type: string title: Webhook Url cost: anyOf: - type: number - type: 'null' title: Cost description: Cost in credits for this request input_mp: anyOf: - type: number - type: 'null' title: Input Mp description: Input megapixels (2 decimal places) output_mp: anyOf: - type: number - type: 'null' title: Output Mp description: Output megapixels (2 decimal places) type: object required: - id - status - webhook_url title: AsyncWebhookResponse HTTPValidationError: properties: detail: items: $ref: '#/components/schemas/ValidationError' type: array title: Detail type: object title: HTTPValidationError Flux3VideoT2VInputs: properties: prompt: type: string title: Prompt description: Free-form prompt describing the video. aspect_ratio: anyOf: - type: string enum: - '21:9' - '2:1' - '16:9' - '4:3' - '1:1' - '3:4' - '9:16' - type: string const: auto title: Aspect Ratio description: >- Output aspect ratio. `auto` lets the harness choose from the prompt and any references. default: auto duration: anyOf: - type: integer maximum: 20 minimum: 5 - type: string const: auto title: Duration description: >- Video duration in seconds (any whole second from 5 to 20), or `auto` to fit the content. default: auto resolution: type: string enum: - hd - fhd title: Resolution description: >- Video resolution class: `hd`, or `fhd` for a higher-resolution result finished by the video upsampler. Exact dimensions vary with the aspect ratio. default: hd version: type: string const: latest title: Version description: >- Endpoint version. `latest` (default) serves the current release; dated pinnable release tags are added here as they are published. default: latest generate_audio: type: boolean title: Generate Audio description: Generate synchronized audio alongside the video. default: true safety_tolerance: type: integer maximum: 4 minimum: 0 title: Safety Tolerance description: >- Tolerance level for input and output harm moderation. Between 0 and 4, with 0 the strictest. Sexual content is limited to level 3 and hate content to level 2 regardless of the requested tolerance. Contextual safety signals can tighten the requested level. Requests with conditioning media are limited to level 2. default: 2 draft: type: boolean title: Draft description: >- Draft mode: generate a fast preview and return a `draft_cache` download URL in the result. Download that bundle and pass it back later to render the full-quality version of the same generation. default: false mode: type: string const: t2v title: Mode additionalProperties: false type: object required: - prompt - mode title: Flux3VideoT2VInputs Flux3VideoI2VInputs: properties: prompt: type: string title: Prompt description: Free-form prompt describing the video. aspect_ratio: anyOf: - type: string enum: - '21:9' - '2:1' - '16:9' - '4:3' - '1:1' - '3:4' - '9:16' - type: string const: auto title: Aspect Ratio description: >- Output aspect ratio. `auto` lets the harness choose from the prompt and any references. default: auto duration: anyOf: - type: integer maximum: 20 minimum: 5 - type: string const: auto title: Duration description: >- Video duration in seconds (any whole second from 5 to 20), or `auto` to fit the content. default: auto resolution: type: string enum: - hd - fhd title: Resolution description: >- Video resolution class: `hd`, or `fhd` for a higher-resolution result finished by the video upsampler. Exact dimensions vary with the aspect ratio. default: hd version: type: string const: latest title: Version description: >- Endpoint version. `latest` (default) serves the current release; dated pinnable release tags are added here as they are published. default: latest generate_audio: type: boolean title: Generate Audio description: Generate synchronized audio alongside the video. default: true safety_tolerance: type: integer maximum: 4 minimum: 0 title: Safety Tolerance description: >- Tolerance level for input and output harm moderation. Between 0 and 4, with 0 the strictest. Sexual content is limited to level 3 and hate content to level 2 regardless of the requested tolerance. Contextual safety signals can tighten the requested level. Requests with conditioning media are limited to level 2. default: 2 draft: type: boolean title: Draft description: >- Draft mode: generate a fast preview and return a `draft_cache` download URL in the result. Download that bundle and pass it back later to render the full-quality version of the same generation. default: false mode: type: string const: i2v title: Mode keyframes: anyOf: - type: string - prefixItems: - type: number - type: string type: array maxItems: 2 minItems: 2 - items: type: string type: array - items: prefixItems: - type: number - type: string type: array maxItems: 2 minItems: 2 type: array title: Keyframes description: >- Your images become frames of the video; each is an http(s) URL or base64, one to ten total. Plain images: one starts the video, two start and end it, with more the first starts it, the last ends it, and the rest fall evenly in between (3 or more need a set `duration`). To control the timing yourself, send `[seconds, image]` pairs in time order, e.g. `[[0, "..."], [3.5, "..."]]`: each image becomes the frame at that second, and with `duration: "auto"` the video runs to the last pair's second, rounded up (20s max). additionalProperties: false type: object required: - prompt - mode - keyframes title: Flux3VideoI2VInputs description: |- Image continuation: your images become frames of the video. One image starts the video; two start and end it; with more, the first starts it, the last ends it, and the rest fall evenly in between. To control the timing yourself, send `[seconds, image]` pairs: each image becomes the frame at that second. Flux3VideoV2VInputs: properties: prompt: type: string title: Prompt description: Free-form prompt describing the video. aspect_ratio: anyOf: - type: string enum: - '21:9' - '2:1' - '16:9' - '4:3' - '1:1' - '3:4' - '9:16' - type: string const: auto title: Aspect Ratio description: >- Output aspect ratio. `auto` lets the harness choose from the prompt and any references. default: auto duration: anyOf: - type: integer maximum: 20 minimum: 5 - type: string const: auto title: Duration description: >- Video duration in seconds (any whole second from 5 to 20), or `auto` to fit the content. default: auto resolution: type: string enum: - hd - fhd title: Resolution description: >- Video resolution class: `hd`, or `fhd` for a higher-resolution result finished by the video upsampler. Exact dimensions vary with the aspect ratio. default: hd version: type: string const: latest title: Version description: >- Endpoint version. `latest` (default) serves the current release; dated pinnable release tags are added here as they are published. default: latest generate_audio: type: boolean title: Generate Audio description: Generate synchronized audio alongside the video. default: true safety_tolerance: type: integer maximum: 4 minimum: 0 title: Safety Tolerance description: >- Tolerance level for input and output harm moderation. Between 0 and 4, with 0 the strictest. Sexual content is limited to level 3 and hate content to level 2 regardless of the requested tolerance. Contextual safety signals can tighten the requested level. Requests with conditioning media are limited to level 2. default: 2 draft: type: boolean title: Draft description: >- Draft mode: generate a fast preview and return a `draft_cache` download URL in the result. Download that bundle and pass it back later to render the full-quality version of the same generation. default: false mode: type: string const: v2v title: Mode start_video: type: string title: Start Video description: >- The video to continue, an http(s) URL or base64 mp4; the generated clip carries on from its final frames. additionalProperties: false type: object required: - prompt - mode - start_video title: Flux3VideoV2VInputs description: |- Video continuation: the generated clip carries on from your video's final frames. Flux3VideoDraftEnhanceInputs: properties: mode: type: string const: draft_enhance title: Mode draft_cache: type: string title: Draft Cache description: >- Encrypted draft-cache bundle from a prior `draft` generation. Primary form: the base64-encoded `.bin` file downloaded from the prior result's `draft_cache` URL. An http(s) URL is also accepted for replays within the download URL's expiry window. The harness is skipped and the original FLUX call is reproduced at full quality; the original inputs are embedded in the bundle. resolution: type: string enum: - hd - fhd title: Resolution description: >- Video resolution class of the enhanced result: `fhd` (default) finishes the reproduced generation with the video upsampler, `hd` returns it without that pass. The reproduced generation is identical either way; only the finishing pass differs. default: fhd safety_tolerance: type: integer maximum: 4 minimum: 0 title: Safety Tolerance description: >- Tolerance level for replay input and output harm moderation. Between 0 and 4, with 0 the strictest. Sexual content is limited to level 3 and hate content to level 2 regardless of the requested tolerance. Bundles containing user media are limited to level 2. default: 2 additionalProperties: false type: object required: - mode - draft_cache title: Flux3VideoDraftEnhanceInputs description: |- Full-quality render of a prior `draft` generation. The bundle pins the generation itself -- the original mode, prompt, seed, and conditioning media -- so `resolution` is the only output choice left to the caller. ValidationError: properties: loc: items: anyOf: - type: string - type: integer type: array title: Location msg: type: string title: Message type: type: string title: Error Type type: object required: - loc - msg - type title: ValidationError securitySchemes: APIKeyHeader: type: apiKey in: header name: x-key ````