# Kling 4.0 Flash Reference to Video > Kling 4.0 Flash generates 720p videos from text prompts and image or video references, with durations from 3 to 20 seconds. ## Overview - **Endpoint**: `https://fal.run/kling/v4/flash/reference-to-video` - **Model ID**: `kling/v4/flash/reference-to-video` - **Category**: video-to-video - **Kind**: inference **Tags**: kling, v4, flash, reference-to-video, audio ## Pricing Video costs **$0.094** per second at **720p**. With video references, it costs **$0.063** per second, billed on the output duration plus the greater of the input video duration and the output duration. For more details, see [fal.ai pricing](https://fal.ai/pricing). ## API Information This model can be used via our HTTP API or more conveniently via our client libraries. See the input and output schema below, as well as the usage examples. ### Input Schema The API accepts the following input parameters: - **`prompt`** (`string`, _required_): Text prompt. Reference inputs as @Image1, @Video1, @Element1, etc. - Examples: "10-second macro product film for the fictional watchmaker AUREL, 16:9, black-velvet studio, one raking strip light. Four shots, hard cuts, slow and precise.\nREFERENCES: @Image1 locks the dial (grand-feu white enamel, slim blued leaf hands, polished baton indices, small seconds at six, 'AUREL' under twelve). @Image2 locks the movement seen through the caseback. @Image3 locks the case flank, brushed and polished bevels and the fluted crown. @Image4 is the end card - reproduce it exactly.\n0-3 s: @Image1. A motion-control slider glides slowly left to right across the dial at macro distance; the strip-light reflection travels across the domed sapphire; the blued small-seconds hand steps in fine six-beats-per-second increments.\n3-6 s: @Image2. Slow push-in on the movement; the balance wheel oscillates steadily (about three swings per second), the hairspring breathing; light rakes across the Geneva stripes and blued screws.\n6-8 s: @Image3. Locked-off extreme macro; the thin strip rim light travels slowly along the mirror-polished bevel and over the fluted crown, one bright glint running edge to edge across the dry steel.\n8-10 s: hard cut as the glint leaves the crown to the end card of @Image4, held perfectly still; 'AUREL' and 'CALIBRE 7' pixel-identical.\nAUDIO: close-mic mechanical ticking at six beats per second, crisp and dominant; a low warm analogue drone under it; a faint high metallic shimmer as the glint crosses the crown at about 7.8 s; one soft low piano note at 8 s that rings under the end card. No voice-over, no music beyond that note.\nLOCKS: same case, dial, hands and indices across shots; no numerals appearing on the dial; no text except the dial print and the end card.\nAVOID: any water, droplets, condensation or liquid anywhere, dial text drift or extra words, warped hands, hands moving at the wrong speed, a ticking quartz-style seconds jump on a sweep dial, lens flares, CG plastic look, fast cutting." - **`duration`** (`integer`, _optional_): Video duration in seconds (3-20). Default value: `5` - Default: `5` - Range: `3` to `20` - Examples: 10 - **`start_image_url`** (`string`, _optional_): Optional image to use as the first frame. Counts toward the image reference limit. - **`image_urls`** (`list`, _optional_): Reference images, referenced as @Image1, @Image2, ... (max 10). - Array of string - Examples: ["https://v3b.fal.media/files/b/0aad8e7f/Rqg-pv2QnjyCj_qyhkb1L_image_1.png","https://v3b.fal.media/files/b/0aad8e7f/tZbpEZhQSFIdaP-4bA7mZ_image_2.png","https://v3b.fal.media/files/b/0aad90a6/fUtkQcN-M0lfbW0d64cgI_image_3.png","https://v3b.fal.media/files/b/0aad8e7f/-qNeGnjPP6Xv0L811MyJg_image_4.png"] - **`video_urls`** (`list`, _optional_): Reference videos (3-20s each, 20s combined with element videos), referenced as @Video1, @Video2, ... (max 5). - Array of string - Examples: null - **`elements`** (`list`, _optional_): Characters/objects, referenced as @Element1, @Element2, ... (max 7). Up to 3 may be video elements. Flash does not accept voices. - Array of KlingV4FlashElement - Examples: null - **`resolution`** (`ResolutionEnum`, _optional_): Output resolution. Flash is 720p only. Default value: `"720p"` - Default: `"720p"` - Options: `"720p"` - **`aspect_ratio`** (`AspectRatioEnum`, _optional_): Aspect ratio of the generated video. 'auto' adapts it to the references. Default value: `"16:9"` - Default: `"16:9"` - Options: `"auto"`, `"16:9"`, `"9:16"`, `"1:1"`, `"21:9"` **Required Parameters Example**: ```json { "prompt": "10-second macro product film for the fictional watchmaker AUREL, 16:9, black-velvet studio, one raking strip light. Four shots, hard cuts, slow and precise.\nREFERENCES: @Image1 locks the dial (grand-feu white enamel, slim blued leaf hands, polished baton indices, small seconds at six, 'AUREL' under twelve). @Image2 locks the movement seen through the caseback. @Image3 locks the case flank, brushed and polished bevels and the fluted crown. @Image4 is the end card - reproduce it exactly.\n0-3 s: @Image1. A motion-control slider glides slowly left to right across the dial at macro distance; the strip-light reflection travels across the domed sapphire; the blued small-seconds hand steps in fine six-beats-per-second increments.\n3-6 s: @Image2. Slow push-in on the movement; the balance wheel oscillates steadily (about three swings per second), the hairspring breathing; light rakes across the Geneva stripes and blued screws.\n6-8 s: @Image3. Locked-off extreme macro; the thin strip rim light travels slowly along the mirror-polished bevel and over the fluted crown, one bright glint running edge to edge across the dry steel.\n8-10 s: hard cut as the glint leaves the crown to the end card of @Image4, held perfectly still; 'AUREL' and 'CALIBRE 7' pixel-identical.\nAUDIO: close-mic mechanical ticking at six beats per second, crisp and dominant; a low warm analogue drone under it; a faint high metallic shimmer as the glint crosses the crown at about 7.8 s; one soft low piano note at 8 s that rings under the end card. No voice-over, no music beyond that note.\nLOCKS: same case, dial, hands and indices across shots; no numerals appearing on the dial; no text except the dial print and the end card.\nAVOID: any water, droplets, condensation or liquid anywhere, dial text drift or extra words, warped hands, hands moving at the wrong speed, a ticking quartz-style seconds jump on a sweep dial, lens flares, CG plastic look, fast cutting." } ``` **Full Example**: ```json { "prompt": "10-second macro product film for the fictional watchmaker AUREL, 16:9, black-velvet studio, one raking strip light. Four shots, hard cuts, slow and precise.\nREFERENCES: @Image1 locks the dial (grand-feu white enamel, slim blued leaf hands, polished baton indices, small seconds at six, 'AUREL' under twelve). @Image2 locks the movement seen through the caseback. @Image3 locks the case flank, brushed and polished bevels and the fluted crown. @Image4 is the end card - reproduce it exactly.\n0-3 s: @Image1. A motion-control slider glides slowly left to right across the dial at macro distance; the strip-light reflection travels across the domed sapphire; the blued small-seconds hand steps in fine six-beats-per-second increments.\n3-6 s: @Image2. Slow push-in on the movement; the balance wheel oscillates steadily (about three swings per second), the hairspring breathing; light rakes across the Geneva stripes and blued screws.\n6-8 s: @Image3. Locked-off extreme macro; the thin strip rim light travels slowly along the mirror-polished bevel and over the fluted crown, one bright glint running edge to edge across the dry steel.\n8-10 s: hard cut as the glint leaves the crown to the end card of @Image4, held perfectly still; 'AUREL' and 'CALIBRE 7' pixel-identical.\nAUDIO: close-mic mechanical ticking at six beats per second, crisp and dominant; a low warm analogue drone under it; a faint high metallic shimmer as the glint crosses the crown at about 7.8 s; one soft low piano note at 8 s that rings under the end card. No voice-over, no music beyond that note.\nLOCKS: same case, dial, hands and indices across shots; no numerals appearing on the dial; no text except the dial print and the end card.\nAVOID: any water, droplets, condensation or liquid anywhere, dial text drift or extra words, warped hands, hands moving at the wrong speed, a ticking quartz-style seconds jump on a sweep dial, lens flares, CG plastic look, fast cutting.", "duration": 10, "image_urls": [ "https://v3b.fal.media/files/b/0aad8e7f/Rqg-pv2QnjyCj_qyhkb1L_image_1.png", "https://v3b.fal.media/files/b/0aad8e7f/tZbpEZhQSFIdaP-4bA7mZ_image_2.png", "https://v3b.fal.media/files/b/0aad90a6/fUtkQcN-M0lfbW0d64cgI_image_3.png", "https://v3b.fal.media/files/b/0aad8e7f/-qNeGnjPP6Xv0L811MyJg_image_4.png" ], "video_urls": null, "elements": null, "resolution": "720p", "aspect_ratio": "16:9" } ``` ### Output Schema The API returns the following output format: - **`video`** (`File`, _required_): The generated video. - Examples: {"url":"https://v3b.fal.media/files/b/0ab11f9f/7HQolYvL9QnowfKbZGQav_r2v.mp4","content_type":"video/mp4","file_size":12186191,"file_name":"output.mp4"} **Example Response**: ```json { "video": { "url": "https://v3b.fal.media/files/b/0ab11f9f/7HQolYvL9QnowfKbZGQav_r2v.mp4", "content_type": "video/mp4", "file_size": 12186191, "file_name": "output.mp4" } } ``` ## Usage Examples ### cURL ```bash curl --request POST \ --url https://fal.run/kling/v4/flash/reference-to-video \ --header "Authorization: Key $FAL_KEY" \ --header "Content-Type: application/json" \ --data '{ "prompt": "10-second macro product film for the fictional watchmaker AUREL, 16:9, black-velvet studio, one raking strip light. Four shots, hard cuts, slow and precise.\nREFERENCES: @Image1 locks the dial (grand-feu white enamel, slim blued leaf hands, polished baton indices, small seconds at six, 'AUREL' under twelve). @Image2 locks the movement seen through the caseback. @Image3 locks the case flank, brushed and polished bevels and the fluted crown. @Image4 is the end card - reproduce it exactly.\n0-3 s: @Image1. A motion-control slider glides slowly left to right across the dial at macro distance; the strip-light reflection travels across the domed sapphire; the blued small-seconds hand steps in fine six-beats-per-second increments.\n3-6 s: @Image2. Slow push-in on the movement; the balance wheel oscillates steadily (about three swings per second), the hairspring breathing; light rakes across the Geneva stripes and blued screws.\n6-8 s: @Image3. Locked-off extreme macro; the thin strip rim light travels slowly along the mirror-polished bevel and over the fluted crown, one bright glint running edge to edge across the dry steel.\n8-10 s: hard cut as the glint leaves the crown to the end card of @Image4, held perfectly still; 'AUREL' and 'CALIBRE 7' pixel-identical.\nAUDIO: close-mic mechanical ticking at six beats per second, crisp and dominant; a low warm analogue drone under it; a faint high metallic shimmer as the glint crosses the crown at about 7.8 s; one soft low piano note at 8 s that rings under the end card. No voice-over, no music beyond that note.\nLOCKS: same case, dial, hands and indices across shots; no numerals appearing on the dial; no text except the dial print and the end card.\nAVOID: any water, droplets, condensation or liquid anywhere, dial text drift or extra words, warped hands, hands moving at the wrong speed, a ticking quartz-style seconds jump on a sweep dial, lens flares, CG plastic look, fast cutting." }' ``` ### Python Ensure you have the Python client installed: ```bash pip install fal-client ``` Then use the API client to make requests: ```python import fal_client def on_queue_update(update): if isinstance(update, fal_client.InProgress): for log in update.logs: print(log["message"]) result = fal_client.subscribe( "kling/v4/flash/reference-to-video", arguments={ "prompt": "10-second macro product film for the fictional watchmaker AUREL, 16:9, black-velvet studio, one raking strip light. Four shots, hard cuts, slow and precise. REFERENCES: @Image1 locks the dial (grand-feu white enamel, slim blued leaf hands, polished baton indices, small seconds at six, 'AUREL' under twelve). @Image2 locks the movement seen through the caseback. @Image3 locks the case flank, brushed and polished bevels and the fluted crown. @Image4 is the end card - reproduce it exactly. 0-3 s: @Image1. A motion-control slider glides slowly left to right across the dial at macro distance; the strip-light reflection travels across the domed sapphire; the blued small-seconds hand steps in fine six-beats-per-second increments. 3-6 s: @Image2. Slow push-in on the movement; the balance wheel oscillates steadily (about three swings per second), the hairspring breathing; light rakes across the Geneva stripes and blued screws. 6-8 s: @Image3. Locked-off extreme macro; the thin strip rim light travels slowly along the mirror-polished bevel and over the fluted crown, one bright glint running edge to edge across the dry steel. 8-10 s: hard cut as the glint leaves the crown to the end card of @Image4, held perfectly still; 'AUREL' and 'CALIBRE 7' pixel-identical. AUDIO: close-mic mechanical ticking at six beats per second, crisp and dominant; a low warm analogue drone under it; a faint high metallic shimmer as the glint crosses the crown at about 7.8 s; one soft low piano note at 8 s that rings under the end card. No voice-over, no music beyond that note. LOCKS: same case, dial, hands and indices across shots; no numerals appearing on the dial; no text except the dial print and the end card. AVOID: any water, droplets, condensation or liquid anywhere, dial text drift or extra words, warped hands, hands moving at the wrong speed, a ticking quartz-style seconds jump on a sweep dial, lens flares, CG plastic look, fast cutting." }, with_logs=True, on_queue_update=on_queue_update, ) print(result) ``` ### JavaScript Ensure you have the JavaScript client installed: ```bash npm install --save @fal-ai/client ``` Then use the API client to make requests: ```javascript import { fal } from "@fal-ai/client"; const result = await fal.subscribe("kling/v4/flash/reference-to-video", { input: { prompt: "10-second macro product film for the fictional watchmaker AUREL, 16:9, black-velvet studio, one raking strip light. Four shots, hard cuts, slow and precise. REFERENCES: @Image1 locks the dial (grand-feu white enamel, slim blued leaf hands, polished baton indices, small seconds at six, 'AUREL' under twelve). @Image2 locks the movement seen through the caseback. @Image3 locks the case flank, brushed and polished bevels and the fluted crown. @Image4 is the end card - reproduce it exactly. 0-3 s: @Image1. A motion-control slider glides slowly left to right across the dial at macro distance; the strip-light reflection travels across the domed sapphire; the blued small-seconds hand steps in fine six-beats-per-second increments. 3-6 s: @Image2. Slow push-in on the movement; the balance wheel oscillates steadily (about three swings per second), the hairspring breathing; light rakes across the Geneva stripes and blued screws. 6-8 s: @Image3. Locked-off extreme macro; the thin strip rim light travels slowly along the mirror-polished bevel and over the fluted crown, one bright glint running edge to edge across the dry steel. 8-10 s: hard cut as the glint leaves the crown to the end card of @Image4, held perfectly still; 'AUREL' and 'CALIBRE 7' pixel-identical. AUDIO: close-mic mechanical ticking at six beats per second, crisp and dominant; a low warm analogue drone under it; a faint high metallic shimmer as the glint crosses the crown at about 7.8 s; one soft low piano note at 8 s that rings under the end card. No voice-over, no music beyond that note. LOCKS: same case, dial, hands and indices across shots; no numerals appearing on the dial; no text except the dial print and the end card. AVOID: any water, droplets, condensation or liquid anywhere, dial text drift or extra words, warped hands, hands moving at the wrong speed, a ticking quartz-style seconds jump on a sweep dial, lens flares, CG plastic look, fast cutting." }, logs: true, onQueueUpdate: (update) => { if (update.status === "IN_PROGRESS") { update.logs.map((log) => log.message).forEach(console.log); } }, }); console.log(result.data); console.log(result.requestId); ``` ## Deploy your own model To run your own weights, a fine-tune or a custom pipeline on the GPUs that serve this API, deploy it on fal Serverless. GPUs from H100 (80GB) to B300 (288GB), billed per second; an H100 is $4.50/h list, as low as $2.49/h. Prices, GPU specs and deploy steps: [fal Serverless](https://fal.ai/serverless.md). ## Additional Resources ### Documentation - [Model Playground](https://fal.ai/models/kling/v4/flash/reference-to-video) - [API Documentation](https://fal.ai/models/kling/v4/flash/reference-to-video/api) - [OpenAPI Schema](https://fal.ai/api/openapi/queue/openapi.json?endpoint_id=kling/v4/flash/reference-to-video) ### fal.ai Platform - [Platform Documentation](https://fal.ai/docs/documentation) - [Python Client](https://fal.ai/docs/api-reference/client-libraries/python) - [JavaScript Client](https://fal.ai/docs/api-reference/client-libraries/javascript) ### Other agent-readable surfaces This file covers one model. To find anything else: - [Platform overview](https://fal.ai/llms.txt): Entry points and representative endpoint IDs - [Documentation index](https://fal.ai/docs/llms.txt): Every documentation page - [Full documentation text](https://fal.ai/docs/llms-full.txt): The whole documentation inlined - [Pricing](https://fal.ai/pricing.md): Model API and Serverless GPU prices - [fal Serverless](https://fal.ai/serverless.md): Deploy your own models: GPU prices, specs and steps - Any other model: `https://fal.ai/models//llms.txt`