FLUX Video Upscale is Black Forest Labs’ FLUX 3-powered video super-resolution workflow. It receives an existing clip and returns an enlarged MP4, with the source framing and audio carried into the result. A short optional description can steer the enhanced visual detail. Its two modes make a meaningful choice between staying close to the supplied appearance and generating more aggressive detail, rather than treating every larger image as an equally faithful reconstruction.
The creativity setting selects the mode. A value of 0 is the precise, source-faithful profile: the manufacturer recommends it when a face, product or brand asset needs to keep its identity. A value of 1 enables creative detail enhancement and is the default. It can restore or invent texture more aggressively, which is useful for generated footage, scenery, crowds or surfaces whose detail can be developed creatively.
The maker explicitly warns that creative enhancement does not preserve faces and products as strictly. Check the feature that must remain identifiable: facial appearance, a product outline, lettering, a material pattern or a logo. A visual that looks sharper can still contain a changed or newly invented detail. For a product demonstration, use the intended product appearance as the review criterion rather than selecting a version only because its texture is more elaborate.
The optional prompt describes the content of the clip to guide enhancement; an empty description requests a neutral upscale. An original generated-scene brief could identify a moss-covered stone wall, soft daylight and fine leaf texture. Keep the wording about the material that should be refined. A request to replace an object, rebuild the setting or change dialogue belongs to an editing decision, while this profile retains the source audio track.
The documented upscale_factor ranges from 1.5 to 3, with 2 as the default. It enlarges the source resolution while preserving the aspect ratio. The selected factor and the input dimensions determine the intended output, so a landscape clip and a portrait composition need to be planned from their actual frames. Prepare the target display size before choosing how much enlargement is useful.
The exact specification caps each output frame at 13.75 of the manufacturer’s megapixel units. Its documentation defines one such unit as 1,048,576 pixels, giving a ceiling of 14,417,920 pixels, approximately 14.4 million. That explains the rounded 14.4-megapixel wording in the guide. A large source can therefore receive less enlargement than the requested factor. Check the delivered dimensions rather than assuming that a 3× request always produces three times each source dimension.
The published source-resolution ceiling is 2560×1440. This workflow increases an eligible source toward 4K and beyond; it does not accept an already larger source simply because a higher output is desired. Downscale or prepare an appropriate working copy before using that documented profile. Keep the original recording separately so the final comparison is not limited to a resized intermediate.
The maker’s ordinary preparation guidance limits sources to twenty seconds and fifty MB. A clip is supplied through a video URL or encoded MP4. Its exact reference notes that processing covers the first twenty seconds: a small length overshoot can be processed only to that mark, while a substantially longer clip is rejected. No numerical tolerance is published. Trim to the intended twenty-second-or-shorter segment instead of depending on that unspecified boundary behavior.
Choose the least compressed version available. Compression artifacts limit the amount of real detail the upscaler can recover, and creative enhancement can make a plausible texture without establishing what the original scene contained. Complete the intended edit and trim before enlarging the selected clip. An original workflow might first finish a product shot, choose the useful action interval and then upscale that final selection, retaining the unmodified source for comparison.
The result keeps the source audio track. Inspect its timing with the picture, especially at the beginning and end of a selected segment. The documented profile does not supply an independent frame-rate selector, new dialogue generation or image-reference editing controls. Its purpose is the larger video and chosen treatment of detail, with the source’s temporal and sonic material retained within the processed segment.
The producer workflow is asynchronous: a submitted task supplies an identity and status location, and its completed result contains a signed MP4 delivery link. That link expires after about one hour. Retain the actual clip, source, selected factor, creativity mode and content description instead of relying on the temporary location. Keep failed or moderated status distinct from a completed video.
Review individual frames for faces, markings and surface changes, then watch the sequence for continuity, motion and audio alignment. Precise and creative outputs answer different visual requirements; the useful artifact is the enlarged clip that meets the chosen fidelity goal at the intended display size. A larger resolution alone is not evidence that the original identity, wording or fine detail remained correct.
", "aboutText": "FLUX Video Upscale is Black Forest Labs’ FLUX 3-powered video super-resolution workflow. It receives an existing clip and returns an enlarged MP4, with the source framing and audio carried into the result. A short optional description can steer the enhanced visual detail. Its two modes make a meaningful choice between staying close to the supplied appearance and generating more aggressive detail, rather than treating every larger image as an equally faithful reconstruction.\n\n### Choose precise or creative enhancement\n\nThe creativity setting selects the mode. A value of 0 is the precise, source-faithful profile: the manufacturer recommends it when a face, product or brand asset needs to keep its identity. A value of 1 enables creative detail enhancement and is the default. It can restore or invent texture more aggressively, which is useful for generated footage, scenery, crowds or surfaces whose detail can be developed creatively.\n\nThe maker explicitly warns that creative enhancement does not preserve faces and products as strictly. Check the feature that must remain identifiable: facial appearance, a product outline, lettering, a material pattern or a logo. A visual that looks sharper can still contain a changed or newly invented detail. For a product demonstration, use the intended product appearance as the review criterion rather than selecting a version only because its texture is more elaborate.\n\nThe optional prompt describes the content of the clip to guide enhancement; an empty description requests a neutral upscale. An original generated-scene brief could identify a moss-covered stone wall, soft daylight and fine leaf texture. Keep the wording about the material that should be refined. A request to replace an object, rebuild the setting or change dialogue belongs to an editing decision, while this profile retains the source audio track.\n\n### Scale and output frame area\n\nThe documented upscale_factor ranges from 1.5 to 3, with 2 as the default. It enlarges the source resolution while preserving the aspect ratio. The selected factor and the input dimensions determine the intended output, so a landscape clip and a portrait composition need to be planned from their actual frames. Prepare the target display size before choosing how much enlargement is useful.\n\nThe exact specification caps each output frame at 13.75 of the manufacturer’s megapixel units. Its documentation defines one such unit as 1,048,576 pixels, giving a ceiling of 14,417,920 pixels, approximately 14.4 million. That explains the rounded 14.4-megapixel wording in the guide. A large source can therefore receive less enlargement than the requested factor. Check the delivered dimensions rather than assuming that a 3× request always produces three times each source dimension.\n\nThe published source-resolution ceiling is 2560×1440. This workflow increases an eligible source toward 4K and beyond; it does not accept an already larger source simply because a higher output is desired. Downscale or prepare an appropriate working copy before using that documented profile. Keep the original recording separately so the final comparison is not limited to a resized intermediate.\n\n### Prepare a short, readable source\n\nThe maker’s ordinary preparation guidance limits sources to twenty seconds and fifty MB. A clip is supplied through a video URL or encoded MP4. Its exact reference notes that processing covers the first twenty seconds: a small length overshoot can be processed only to that mark, while a substantially longer clip is rejected. No numerical tolerance is published. Trim to the intended twenty-second-or-shorter segment instead of depending on that unspecified boundary behavior.\n\nChoose the least compressed version available. Compression artifacts limit the amount of real detail the upscaler can recover, and creative enhancement can make a plausible texture without establishing what the original scene contained. Complete the intended edit and trim before enlarging the selected clip. An original workflow might first finish a product shot, choose the useful action interval and then upscale that final selection, retaining the unmodified source for comparison.\n\nThe result keeps the source audio track. Inspect its timing with the picture, especially at the beginning and end of a selected segment. The documented profile does not supply an independent frame-rate selector, new dialogue generation or image-reference editing controls. Its purpose is the larger video and chosen treatment of detail, with the source’s temporal and sonic material retained within the processed segment.\n\n### Keep and check the delivered clip\n\nThe producer workflow is asynchronous: a submitted task supplies an identity and status location, and its completed result contains a signed MP4 delivery link. That link expires after about one hour. Retain the actual clip, source, selected factor, creativity mode and content description instead of relying on the temporary location. Keep failed or moderated status distinct from a completed video.\n\nReview individual frames for faces, markings and surface changes, then watch the sequence for continuity, motion and audio alignment. Precise and creative outputs answer different visual requirements; the useful artifact is the enlarged clip that meets the chosen fidelity goal at the intended display size. A larger resolution alone is not evidence that the original identity, wording or fine detail remained correct.", "editorialOverlay": { "copy_fields": null, "copy_sha256": null, "copy_source": null, "html_sha256": "458b668d9115b908b572c40795c0e2bc391ee4cd424dac01d29cdcb8f5305a30", "module_sha256": "ee9dc1c50b1e6a8fb7fa062030fcb4646b9aaaeb08be513b723a5a92a7290f06", "raw_model_unchanged": true, "schema": "ailoot-native-model-editorial-v1" }, "locale": "en", "methods": [ { "html": "black-forest-labs/flux-video-upscaleFLUX Video Upscale enlarges an existing clip through source-faithful or creative detail enhancement. Its documented profile offers 1.5×–3× scaling, preserves the source audio and framing, and limits source length, input resolution and output frame area.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
FLUX Video Upscale is Black Forest Labs’ FLUX 3-powered video super-resolution workflow. It receives an existing clip and returns an enlarged MP4, with the source framing and audio carried into the result. A short optional description can steer the enhanced visual detail. Its two modes make a meaningful choice between staying close to the supplied appearance and generating more aggressive detail, rather than treating every larger image as an equally faithful reconstruction.
The creativity setting selects the mode. A value of 0 is the precise, source-faithful profile: the manufacturer recommends it when a face, product or brand asset needs to keep its identity. A value of 1 enables creative detail enhancement and is the default. It can restore or invent texture more aggressively, which is useful for generated footage, scenery, crowds or surfaces whose detail can be developed creatively.
The maker explicitly warns that creative enhancement does not preserve faces and products as strictly. Check the feature that must remain identifiable: facial appearance, a product outline, lettering, a material pattern or a logo. A visual that looks sharper can still contain a changed or newly invented detail. For a product demonstration, use the intended product appearance as the review criterion rather than selecting a version only because its texture is more elaborate.
The optional prompt describes the content of the clip to guide enhancement; an empty description requests a neutral upscale. An original generated-scene brief could identify a moss-covered stone wall, soft daylight and fine leaf texture. Keep the wording about the material that should be refined. A request to replace an object, rebuild the setting or change dialogue belongs to an editing decision, while this profile retains the source audio track.
The documented upscale_factor ranges from 1.5 to 3, with 2 as the default. It enlarges the source resolution while preserving the aspect ratio. The selected factor and the input dimensions determine the intended output, so a landscape clip and a portrait composition need to be planned from their actual frames. Prepare the target display size before choosing how much enlargement is useful.
The exact specification caps each output frame at 13.75 of the manufacturer’s megapixel units. Its documentation defines one such unit as 1,048,576 pixels, giving a ceiling of 14,417,920 pixels, approximately 14.4 million. That explains the rounded 14.4-megapixel wording in the guide. A large source can therefore receive less enlargement than the requested factor. Check the delivered dimensions rather than assuming that a 3× request always produces three times each source dimension.
The published source-resolution ceiling is 2560×1440. This workflow increases an eligible source toward 4K and beyond; it does not accept an already larger source simply because a higher output is desired. Downscale or prepare an appropriate working copy before using that documented profile. Keep the original recording separately so the final comparison is not limited to a resized intermediate.
The maker’s ordinary preparation guidance limits sources to twenty seconds and fifty MB. A clip is supplied through a video URL or encoded MP4. Its exact reference notes that processing covers the first twenty seconds: a small length overshoot can be processed only to that mark, while a substantially longer clip is rejected. No numerical tolerance is published. Trim to the intended twenty-second-or-shorter segment instead of depending on that unspecified boundary behavior.
Choose the least compressed version available. Compression artifacts limit the amount of real detail the upscaler can recover, and creative enhancement can make a plausible texture without establishing what the original scene contained. Complete the intended edit and trim before enlarging the selected clip. An original workflow might first finish a product shot, choose the useful action interval and then upscale that final selection, retaining the unmodified source for comparison.
The result keeps the source audio track. Inspect its timing with the picture, especially at the beginning and end of a selected segment. The documented profile does not supply an independent frame-rate selector, new dialogue generation or image-reference editing controls. Its purpose is the larger video and chosen treatment of detail, with the source’s temporal and sonic material retained within the processed segment.
The producer workflow is asynchronous: a submitted task supplies an identity and status location, and its completed result contains a signed MP4 delivery link. That link expires after about one hour. Retain the actual clip, source, selected factor, creativity mode and content description instead of relying on the temporary location. Keep failed or moderated status distinct from a completed video.
Review individual frames for faces, markings and surface changes, then watch the sequence for continuity, motion and audio alignment. Precise and creative outputs answer different visual requirements; the useful artifact is the enlarged clip that meets the chosen fidelity goal at the intended display size. A larger resolution alone is not evidence that the original identity, wording or fine detail remained correct.
POST /v1/videostext, image, audio, videovideo| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits black-forest-labs/flux-video-upscale | Default— |
Fieldinput_references | Typearray | RequirementRequired | Values & limits Minimum items: 1 Maximum items: 1 | Default— |
Fieldupscale_factor | Typenumber | RequirementRequired | Values & limits Minimum: 1.5 Maximum: 3 | Default— |
Fieldcreativity | Typeinteger | RequirementRequired | Values & limits 0, 1 | Default— |
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, input_references, upscale_factor, creativity |
Parametermaterial_types | Contracttext, image, audio, video |
Parametermaterial_conditions | ContractUse exactly one actual completed owned video asset content URL with its returned signature. The asset owner, type and stored byte witness must match the request. |
Lifecycle: asynchronous_job_then_owned_outputs
GET /v1/videos/{request_id} · actual returned id or polling_url
GET /v1/videos/{request_id}/content
The displayed rates estimate the cost before generation. The final charge is calculated from the cost recorded for the original generation. Measured output duration and forecast quantities are not the final billable quantity.
Tariff references do not add request options. Supported parameters are listed for the selected method.
Match each charge to the conditions shown. Use one rate for each set of alternative configurations.
| Configuration | Rate schedule | Forecast quantity | Billing unit | Billing conditions | Pixels per megapixel | Charge |
|---|---|---|---|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Forecast quantity— | Billing unit— | Billing conditions— | Pixels per megapixel— | Charge Input · Input tokens · Estimate for comparison only — / 1K tokens Output · Output tokens · Estimate for comparison only — / 1K tokens Input · Cache read · Estimate for comparison only — / 1K tokens Input · Cache write · Estimate for comparison only — / 1K tokens Input · Cache write, 1 hour · Estimate for comparison only — / 1K tokens |
| ConfigurationVideo price | Rate schedule— | Forecast quantity— | Billing unitper megapixel-second | Billing conditionsquality: creative | Pixels per megapixel1,048,576 | Charge Output · Video price · Estimate for comparison only $0.21 per megapixel-second |
| ConfigurationVideo price | Rate schedule— | Forecast quantity— | Billing unitper megapixel-second | Billing conditionsquality: precise | Pixels per megapixel1,048,576 | Charge Output · Video price · Estimate for comparison only $0.15 per megapixel-second |
| ConfigurationEstimated generation cost | Rate schedule— | Forecast quantityone generation estimate | Billing unit— | Billing conditions— | Pixels per megapixel— | Charge Output · Estimated generation cost · Estimate for comparison only $0.15 estimate before generation This estimate is not an additional charge or a fixed rate for the completed generation. |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\n: "${AILOOT_INPUT_REFERENCES_JSON:?Set the actual documented input value}"\nexport AILOOT_INPUT_REFERENCES_JSON\n: "${AILOOT_UPSCALE_FACTOR:?Set the actual documented input value}"\nexport AILOOT_UPSCALE_FACTOR\n: "${AILOOT_CREATIVITY:?Set the actual documented input value}"\nexport AILOOT_CREATIVITY\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\nrequest={'model': 'black-forest-labs/flux-video-upscale', 'input_references': '${ACTUAL_OWNED_INPUT_REFERENCES}', 'upscale_factor': '${ACTUAL_UPSCALE_FACTOR}', 'creativity': '${ACTUAL_CREATIVITY}'}\nrequest['input_references']=json.loads(os.environ['AILOOT_INPUT_REFERENCES_JSON'])\nrequest['upscale_factor']=float(os.environ['AILOOT_UPSCALE_FACTOR'])\nrequest['creativity']=int(os.environ['AILOOT_CREATIVITY'])\nPath('request.json').write_text(json.dumps(request,ensure_ascii=False),encoding='utf-8')\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/videos \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o response.bin\n# Retain the complete response and headers, every actual output, original returned identifiers and all usage fields. Recover the original operation without submitting creation again.\nGET /v1/videos/{request_id}GET /v1/videos/{request_id}/contentset -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\n: "${AILOOT_STATUS_FILE:?Set the actual completed status JSON file}"\nexport AILOOT_STATUS_FILE\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\nfrom urllib.parse import urlsplit,urljoin\nf=Path(os.environ["AILOOT_STATUS_FILE"])\nif not f.is_file():raise SystemExit("Choose the actual completed status file")\nr=json.loads(f.read_text())\nif r.get("status")!="completed" or not isinstance(r.get("data"),list) or not r["data"]:raise SystemExit("Wait for a completed status with asset paths")\nurls=[]\nfor x in r["data"]:\n if not isinstance(x,str) or "\\n" in x or "\\r" in x:raise SystemExit("Invalid asset path")\n u=urljoin("https://api.ailoot.ai",x)\n z=urlsplit(u)\n if z.scheme!="https" or z.netloc!="api.ailoot.ai" or z.username or z.password:raise SystemExit("Use the actual Ailoot asset paths from status")\n urls.append(u)\nPath("asset-urls.txt").write_text("\\n".join(urls)+"\\n")\nPY_REQUEST\nindex=0\nwhile IFS= read -r asset_url; do\n curl --fail-with-body --silent --show-error "$asset_url" \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -o "video-${index}.mp4"\n index=$((index + 1))\ndone < asset-urls.txt\nSave every completed video or returned file; retrieve every actual returned asset URL.
Retain the complete response and headers, every actual output, original returned identifiers and all usage fields. Recover the original operation without submitting creation again.
Keep X-Client-Request-ID, X-Ailoot-Task-Id and actual returned id/polling_url. Continue the original job; do not submit creation again.
| Field | Type |
|---|---|
Fieldid | Type |
Fieldstatus | Type |
Fielddata[] | Type |
Fieldoutputs[] | Type |
Ailoot Ultra black-forest-labs/flux-video-upscale
Use the method paths shown above.