Krea 2 Large is the larger of Krea’s two original Krea 2 image models, trained from scratch for a broad visual range. Krea describes Large as having more than twice Medium’s size and softer post-training: it retains a rawer, more textured character and offers more flexibility in photographic styles. Its intended strengths include images with film grain, motion blur and a deliberately limited dynamic range, alongside cleaner photographic work.
Use it for editorial photography concepts, product scenes, campaign art direction and visual development where the character of an image matters as much as its subject. A useful brief specifies the subject, setting, framing, lighting, materials and photographic treatment. For example: “A brushed-brass desk lamp beside a closed linen notebook, low evening sunlight, close three-quarter view, subtle grain, restrained warm tones.” This is an image brief; the output is a still image rather than a video sequence.
The generation request requires a text prompt, aspect ratio and resolution. An optional source image enables image-to-image generation: the model starts from that image rather than pure noise. The separate denoising strength runs from 0 to 1; 0 preserves the input and 1 fully replaces it, while the documented default is 0.99. Prepare the source image around the composition and content you want to carry forward, then describe the intended transformation. This control is separate from the influence of a style reference.
Style references guide appearance through supplied images. The model’s request schema accepts up to ten references, each with a strength from 0 to 1 and a default of 0.5. Choose references with a clear visual purpose—such as the lighting of one photograph or the surface treatment of another—and avoid mixing incompatible directions without an explicit hierarchy. References can be supplied through image locations, encoded image data or previously uploaded assets. The model-specific schema limits source-image and style-reference URI strings to 1,024 characters.
A moodboard groups a coherent visual direction into one reusable reference. Krea’s workflow creates the moodboard first and then refers to its existing identifier in generation; a generation request accepts one moodboard. Its schema uses influence from 0 to 1, with 0.23 as the default. Trained styles, typically LoRAs, form another control: each style has an identifier and a strength from -2 to 2. Image references, moodboards and trained styles are distinct inputs, so prepare each for its intended role.
Creativity governs prompt expansion. The raw setting disables expansion; low, medium and high progressively change how much interpretation is added to the written brief. The model-specific request schema defaults to low. Use raw when the wording itself must carry the complete direction, or increase interpretation when exploring a looser concept.
Intensity, Complexity and Movement are independent integer controls from -100 to 100, with 0 as neutral. Lower intensity produces a more muted treatment; higher intensity pushes a bolder style. Lower complexity favors simpler compositions, while higher complexity adds density. Movement changes visual energy, pose and implied motion. It can suggest movement or blur inside a still image; it does not turn the request into video generation. These controls do not change the creativity setting’s prompt-expansion behavior.
For a controlled comparison, hold the brief, reference set and optional seed steady and change one visual control at a time. A practical sequence is to establish composition, select the photographic treatment, adjust texture or energy, and then review the subject’s details. Treat the seed as a recorded generation input alongside the complete settings rather than as a substitute for preserving the finished image.
The model-specific generation schema offers a 1K resolution scale and nine aspect ratios: 1:1, 4:3, 3:2, 16:9, 2.35:1, 4:5, 3:4, 2:3 and 9:16. Select the ratio for the intended placement before generation—for example, a wide campaign composition or a vertical product story. The 1K scale is a generation setting; an enhanced export is a separate output operation.
Krea’s documented generation flow returns a job that may still be pending. Retrieve its status until it completes, then collect the resulting image locations. Keep the final image with its prompt, references and settings so a selected direction can be developed further. Review fine details, visible lettering, object geometry and the match to the brief before using a result in a finished layout.
Large’s photographic emphasis is an edition characteristic, not a guarantee that every prompt will produce a particular aesthetic. Style influence does not provide an exact subject-identity or layout lock, and a strong image-to-image transformation can replace much of the source. For a production asset, make the required composition and text explicit, compare the completed result with the brief, and perform final layout or lettering corrections in the appropriate editor.
", "aboutText": "### Expressive photorealism\n\nKrea 2 Large is the larger of Krea’s two original Krea 2 image models, trained from scratch for a broad visual range. Krea describes Large as having more than twice Medium’s size and softer post-training: it retains a rawer, more textured character and offers more flexibility in photographic styles. Its intended strengths include images with film grain, motion blur and a deliberately limited dynamic range, alongside cleaner photographic work.\n\nUse it for editorial photography concepts, product scenes, campaign art direction and visual development where the character of an image matters as much as its subject. A useful brief specifies the subject, setting, framing, lighting, materials and photographic treatment. For example: “A brushed-brass desk lamp beside a closed linen notebook, low evening sunlight, close three-quarter view, subtle grain, restrained warm tones.” This is an image brief; the output is a still image rather than a video sequence.\n\n### Text, source images and visual references\n\nThe generation request requires a text prompt, aspect ratio and resolution. An optional source image enables image-to-image generation: the model starts from that image rather than pure noise. The separate denoising strength runs from 0 to 1; 0 preserves the input and 1 fully replaces it, while the documented default is 0.99. Prepare the source image around the composition and content you want to carry forward, then describe the intended transformation. This control is separate from the influence of a style reference.\n\nStyle references guide appearance through supplied images. The model’s request schema accepts up to ten references, each with a strength from 0 to 1 and a default of 0.5. Choose references with a clear visual purpose—such as the lighting of one photograph or the surface treatment of another—and avoid mixing incompatible directions without an explicit hierarchy. References can be supplied through image locations, encoded image data or previously uploaded assets. The model-specific schema limits source-image and style-reference URI strings to 1,024 characters.\n\nA moodboard groups a coherent visual direction into one reusable reference. Krea’s workflow creates the moodboard first and then refers to its existing identifier in generation; a generation request accepts one moodboard. Its schema uses influence from 0 to 1, with 0.23 as the default. Trained styles, typically LoRAs, form another control: each style has an identifier and a strength from -2 to 2. Image references, moodboards and trained styles are distinct inputs, so prepare each for its intended role.\n\n### Prompt interpretation and visual controls\n\nCreativity governs prompt expansion. The raw setting disables expansion; low, medium and high progressively change how much interpretation is added to the written brief. The model-specific request schema defaults to low. Use raw when the wording itself must carry the complete direction, or increase interpretation when exploring a looser concept.\n\nIntensity, Complexity and Movement are independent integer controls from -100 to 100, with 0 as neutral. Lower intensity produces a more muted treatment; higher intensity pushes a bolder style. Lower complexity favors simpler compositions, while higher complexity adds density. Movement changes visual energy, pose and implied motion. It can suggest movement or blur inside a still image; it does not turn the request into video generation. These controls do not change the creativity setting’s prompt-expansion behavior.\n\nFor a controlled comparison, hold the brief, reference set and optional seed steady and change one visual control at a time. A practical sequence is to establish composition, select the photographic treatment, adjust texture or energy, and then review the subject’s details. Treat the seed as a recorded generation input alongside the complete settings rather than as a substitute for preserving the finished image.\n\n### Frame choices and completed results\n\nThe model-specific generation schema offers a 1K resolution scale and nine aspect ratios: 1:1, 4:3, 3:2, 16:9, 2.35:1, 4:5, 3:4, 2:3 and 9:16. Select the ratio for the intended placement before generation—for example, a wide campaign composition or a vertical product story. The 1K scale is a generation setting; an enhanced export is a separate output operation.\n\nKrea’s documented generation flow returns a job that may still be pending. Retrieve its status until it completes, then collect the resulting image locations. Keep the final image with its prompt, references and settings so a selected direction can be developed further. Review fine details, visible lettering, object geometry and the match to the brief before using a result in a finished layout.\n\nLarge’s photographic emphasis is an edition characteristic, not a guarantee that every prompt will produce a particular aesthetic. Style influence does not provide an exact subject-identity or layout lock, and a strong image-to-image transformation can replace much of the source. For a production asset, make the required composition and text explicit, compare the completed result with the brief, and perform final layout or lettering corrections in the appropriate editor.", "editorialOverlay": { "copy_fields": null, "copy_sha256": null, "copy_source": null, "html_sha256": "cbbb78cb33ad94c0ed3aee09e84859678558b05005b1fe1bb339e6890584f1ee", "module_sha256": "ee9dc1c50b1e6a8fb7fa062030fcb4646b9aaaeb08be513b723a5a92a7290f06", "raw_model_unchanged": true, "schema": "ailoot-native-model-editorial-v1" }, "locale": "en", "methods": [ { "html": "krea/krea-2-largeKrea 2 Large creates expressive photorealistic images from text, with optional source images, style references and moodboards. Its softer post-training preserves texture and visual character, while separate controls shape prompt interpretation, detail and implied motion.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Krea 2 Large is the larger of Krea’s two original Krea 2 image models, trained from scratch for a broad visual range. Krea describes Large as having more than twice Medium’s size and softer post-training: it retains a rawer, more textured character and offers more flexibility in photographic styles. Its intended strengths include images with film grain, motion blur and a deliberately limited dynamic range, alongside cleaner photographic work.
Use it for editorial photography concepts, product scenes, campaign art direction and visual development where the character of an image matters as much as its subject. A useful brief specifies the subject, setting, framing, lighting, materials and photographic treatment. For example: “A brushed-brass desk lamp beside a closed linen notebook, low evening sunlight, close three-quarter view, subtle grain, restrained warm tones.” This is an image brief; the output is a still image rather than a video sequence.
The generation request requires a text prompt, aspect ratio and resolution. An optional source image enables image-to-image generation: the model starts from that image rather than pure noise. The separate denoising strength runs from 0 to 1; 0 preserves the input and 1 fully replaces it, while the documented default is 0.99. Prepare the source image around the composition and content you want to carry forward, then describe the intended transformation. This control is separate from the influence of a style reference.
Style references guide appearance through supplied images. The model’s request schema accepts up to ten references, each with a strength from 0 to 1 and a default of 0.5. Choose references with a clear visual purpose—such as the lighting of one photograph or the surface treatment of another—and avoid mixing incompatible directions without an explicit hierarchy. References can be supplied through image locations, encoded image data or previously uploaded assets. The model-specific schema limits source-image and style-reference URI strings to 1,024 characters.
A moodboard groups a coherent visual direction into one reusable reference. Krea’s workflow creates the moodboard first and then refers to its existing identifier in generation; a generation request accepts one moodboard. Its schema uses influence from 0 to 1, with 0.23 as the default. Trained styles, typically LoRAs, form another control: each style has an identifier and a strength from -2 to 2. Image references, moodboards and trained styles are distinct inputs, so prepare each for its intended role.
Creativity governs prompt expansion. The raw setting disables expansion; low, medium and high progressively change how much interpretation is added to the written brief. The model-specific request schema defaults to low. Use raw when the wording itself must carry the complete direction, or increase interpretation when exploring a looser concept.
Intensity, Complexity and Movement are independent integer controls from -100 to 100, with 0 as neutral. Lower intensity produces a more muted treatment; higher intensity pushes a bolder style. Lower complexity favors simpler compositions, while higher complexity adds density. Movement changes visual energy, pose and implied motion. It can suggest movement or blur inside a still image; it does not turn the request into video generation. These controls do not change the creativity setting’s prompt-expansion behavior.
For a controlled comparison, hold the brief, reference set and optional seed steady and change one visual control at a time. A practical sequence is to establish composition, select the photographic treatment, adjust texture or energy, and then review the subject’s details. Treat the seed as a recorded generation input alongside the complete settings rather than as a substitute for preserving the finished image.
The model-specific generation schema offers a 1K resolution scale and nine aspect ratios: 1:1, 4:3, 3:2, 16:9, 2.35:1, 4:5, 3:4, 2:3 and 9:16. Select the ratio for the intended placement before generation—for example, a wide campaign composition or a vertical product story. The 1K scale is a generation setting; an enhanced export is a separate output operation.
Krea’s documented generation flow returns a job that may still be pending. Retrieve its status until it completes, then collect the resulting image locations. Keep the final image with its prompt, references and settings so a selected direction can be developed further. Review fine details, visible lettering, object geometry and the match to the brief before using a result in a finished layout.
Large’s photographic emphasis is an edition characteristic, not a guarantee that every prompt will produce a particular aesthetic. Style influence does not provide an exact subject-identity or layout lock, and a strong image-to-image transformation can replace much of the source. For a production asset, make the required composition and text explicit, compare the completed result with the brief, and perform final layout or lettering corrections in the appropriate editor.
POST /v1/images/generationstext, imageimage| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits krea/krea-2-large | Default— |
Fieldprompt | Typestring | RequirementRequired | Values & limits Minimum length: 1 | Default— |
Fieldresolution | Typestring | RequirementOptional | Values & limits 1K | Default— |
Fieldaspect_ratio | Typestring | RequirementOptional | Values & limits 1:1, 4:3, 3:2, 16:9, 4:5, 2:3, 9:16 | Default— |
Fieldinput_references | Typearray | RequirementOptional | Values & limits Minimum items: 0 Maximum items: 1 | Default— |
Fieldprovider | Typeobject | RequirementOptional | Values & limits | Default— |
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, prompt |
Parametermaterial_types | Contracttext, image |
Lifecycle: synchronous_result_then_owned_library
GET /v1/media/tasks/{task_id} · actual returned task_id or media library record
GET /v1/media/assets/{asset_id}/content
GET /v1/media/assets
GET /v1/media/assets/{asset_id}
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 | Charge |
|---|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Forecast quantity— | 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 |
| ConfigurationEstimated generation cost | Rate schedule— | Forecast quantityone generation estimate | Charge Output · Estimated generation cost · Estimate for comparison only $0.12 estimate before generation This estimate is not an additional charge or a fixed rate for the completed generation. |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\nexport AILOOT_API_KEY\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np={'model': 'krea/krea-2-large', 'prompt': 'a red panda astronaut floating in space, studio lighting'}\n\nPath("request.json").write_text(json.dumps(p,ensure_ascii=False))\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/images/generations \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport base64, json, os, subprocess\nfrom pathlib import Path\nfrom urllib.parse import urlsplit, urljoin\n\ndef save_images(items):\n if not items: raise SystemExit("No completed images; retain the full response")\n for index, item in enumerate(items):\n encoded=item.get("b64_json")\n uri=item.get("url", "")\n if not encoded and uri.startswith("data:image/"):\n header, separator, encoded=uri.partition(",")\n if not separator or ";base64" not in header: raise SystemExit("Unsupported image data URI")\n if encoded:\n raw=base64.b64decode(encoded,validate=True)\n elif uri:\n uri=urljoin("https://api.ailoot.ai",uri)\n parsed=urlsplit(uri)\n if parsed.scheme!="https" or not parsed.netloc or parsed.username or parsed.password:\n raise SystemExit("Invalid returned image URL")\n owned=parsed.netloc=="api.ailoot.ai" and parsed.path.startswith("/v1/media/assets/") and parsed.path.endswith("/content")\n args=["curl","--fail-with-body","--silent","--show-error",uri]\n if owned: args += ["-H","Authorization: Bearer "+os.environ["AILOOT_API_KEY"]]\n temporary=Path(f"image-{index}.bin")\n subprocess.run(args+["-o",str(temporary)],check=True)\n raw=temporary.read_bytes()\n temporary.unlink()\n else: raise SystemExit("Image has neither bytes nor URL")\n if not raw: raise SystemExit("Empty image result")\n ext="png" if raw.startswith(b"\\x89PNG") else "jpg" if raw.startswith(b"\\xff\\xd8") else "webp" if raw[:4]==b"RIFF" and raw[8:12]==b"WEBP" else "svg" if raw.lstrip().startswith((b"<svg",b"<?xml")) else "bin"\n Path(f"image-{index}.{ext}").write_bytes(raw)\n print("Every returned image saved; complete response metadata retained")\nsave_images(json.loads(Path("result.json").read_text())["data"])\nPY_RESULT\nDownload each returned image or save its Base64 contents
GET /v1/media/tasks/{task_id}GET /v1/media/assets/{asset_id}/contentGET /v1/media/assetsGET /v1/media/assets/{asset_id}set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/media/assets \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -o owned-assets.json\n# Preserve every returned asset and follow its exact signed content URL. Do not construct a signed token.\nSave every completed image; retrieve each returned URL or decode its Base64 contents.
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 the returned X-Client-Request-ID for reconciliation. Read only actual returned task_id/asset_id values or the owned media library; never build a task URL from the operation ID and never resubmit creation.
| Field | Type |
|---|---|
Fieldcreated | Type |
Fielddata[].url | Type |
Fielddata[].b64_json | Type |
Fieldusage | Type |
Ailoot Ultra krea/krea-2-large
/v1/images