Sora 2 is OpenAI’s video-and-audio generation model for dynamic clips created from a written scene description or an image reference. It generates the visual sequence together with synchronized speech, effects and background sound. The standard Sora 2 profile is positioned for visual exploration, concept development, prototypes and rough cuts. Sora 2 Pro is a separate edition aimed at greater visual fidelity and higher-resolution finishing.
The maker’s prompting guide organizes a useful brief around framing, subject, action, setting and light. Give the scene a coherent visual style, then describe the movement in clear beats. An original product concept might show a small mechanical object turning once on a workbench while the camera moves slowly closer, with soft window light and a subtle motor sound. State what the viewer should notice and how the action ends.
Keep each shot’s camera movement and subject action understandable. A complicated sequence can be divided into distinct shot blocks, each with its own framing, action and lighting. Preserve the same subject details and palette when several clips are intended to form one sequence. The guide emphasizes that shorter shots generally follow direction more reliably and that repeated generation of the same prompt produces different takes.
The model’s launch describes improved physical behavior and continuity across shots, including realistic, cinematic and anime styles. These are intended capabilities to direct and inspect. Review motion, contact between objects, proportions and continuity rather than assuming that a plausible first frame guarantees a coherent sequence. If an action is unclear, simplify it and give it a visible beginning and ending.
Specify spoken lines separately from the visual description and label the speakers consistently. Keep dialogue short enough to fit the action and the intended clip length. For a two-character animation, alternate a brief question and answer rather than asking a short shot to accommodate a long speech. The guide warns that complex, lengthy dialogue can disrupt pacing and synchronization.
Describe the soundscape that belongs to the scene: a small movement’s sound, room ambience or the rhythm of an environmental effect. Explain which sound should coincide with the action. Listen to the resulting speech and effects together with the picture, checking speaker assignment, wording and timing. Synchronized audio is part of the generated clip rather than a separate text answer.
The standard model’s documented output sizes are 720×1280 portrait and 1280×720 landscape. Higher-resolution sizes in the Pro profile should not be attached to standard Sora 2. Size and duration are explicit request settings; writing a different size or asking for a longer film in the prompt does not change those settings. The creation reference lists 4-, 8- and 12-second clip settings, so choose a short, defined beat when preparing a shot. The expanded producer workflow also documents 16- and 20-second clip settings.
An image reference anchors the first frame, and the text describes what happens next. The producer supports JPEG, PNG and WebP references and requires the image to match the target video size. Prepare the subject, environment and composition in that opening image, then ask for a specific movement. This is an opening-frame reference rather than a guarantee of matching every detail throughout the clip.
The expanded producer workflow can create a reusable character asset from a short reference video of an animal, mascot or object. Its preparation guidance calls for an MP4 clip around two to four seconds, 720p to 1080p, in 16:9 or 9:16. Matching the source and output aspect ratio helps avoid a stretched appearance. The generation can reference up to two characters.
Retain the character identifier and mention its name in the prompt; passing the identifier alone is not the guide’s complete preparation. Character assets can accompany an image-guided generation, but video extensions do not accept them. The producer blocks human-likeness character uploads by default and rejects reference images containing human faces in the documented API flow. The Sora app’s personal-character features are a separate product workflow.
An extension takes a completed source video and a continuation prompt, using the full source clip as context for motion, camera direction and scene continuity. The guide describes additions of up to 20 seconds per extension. Its documented extension input is the source video and text; additional character or image references are excluded from that mode. Describe how the existing action continues instead of introducing an unrelated scene without a transition.
Video editing targets a clear change to an existing clip. A focused revision can alter an object’s color or the scene’s palette while retaining the rest of the shot. The producer guide recommends one well-defined change at a time and distinguishes editing a generated video from uploading arbitrary footage, which has a separate eligibility condition. Keep the completed source and identify the exact revision needed before making another edit.
The documented producer video flow is asynchronous: a submitted job can be queued or in progress before it completes or fails. A completed job yields an MP4 video; thumbnail and sprite-sheet assets support previews and browsing. Keep the job identity associated with the prompt, references and returned assets. An initial job response is not the final playable video.
The ordinary download workflow describes a one-hour download-URL lifetime. The expanded batch workflow prepares independent JSON render requests with stable item identifiers, uploaded references and a result mapped back to each shot; its completed assets have a separate download period of up to 24 hours. Preserve the finished clip promptly and inspect the actual picture and sound before treating it as a selected take.
The producer’s documented API content restrictions require material suitable for audiences under 18 and reject copyrighted characters or music and generation of real people. Prepare the scene and references accordingly. The direct model result is video with audio, not generated still images, a conversational tool response or an unlimited film. Retain the exact standard-model identity and the accepted source clip when continuing a visual concept.
", "aboutText": "Sora 2 is OpenAI’s video-and-audio generation model for dynamic clips created from a written scene description or an image reference. It generates the visual sequence together with synchronized speech, effects and background sound. The standard Sora 2 profile is positioned for visual exploration, concept development, prototypes and rough cuts. Sora 2 Pro is a separate edition aimed at greater visual fidelity and higher-resolution finishing.\n\n### Describe a shot that can be seen and heard\n\nThe maker’s prompting guide organizes a useful brief around framing, subject, action, setting and light. Give the scene a coherent visual style, then describe the movement in clear beats. An original product concept might show a small mechanical object turning once on a workbench while the camera moves slowly closer, with soft window light and a subtle motor sound. State what the viewer should notice and how the action ends.\n\nKeep each shot’s camera movement and subject action understandable. A complicated sequence can be divided into distinct shot blocks, each with its own framing, action and lighting. Preserve the same subject details and palette when several clips are intended to form one sequence. The guide emphasizes that shorter shots generally follow direction more reliably and that repeated generation of the same prompt produces different takes.\n\nThe model’s launch describes improved physical behavior and continuity across shots, including realistic, cinematic and anime styles. These are intended capabilities to direct and inspect. Review motion, contact between objects, proportions and continuity rather than assuming that a plausible first frame guarantees a coherent sequence. If an action is unclear, simplify it and give it a visible beginning and ending.\n\n### Direct dialogue and sound separately\n\nSpecify spoken lines separately from the visual description and label the speakers consistently. Keep dialogue short enough to fit the action and the intended clip length. For a two-character animation, alternate a brief question and answer rather than asking a short shot to accommodate a long speech. The guide warns that complex, lengthy dialogue can disrupt pacing and synchronization.\n\nDescribe the soundscape that belongs to the scene: a small movement’s sound, room ambience or the rhythm of an environmental effect. Explain which sound should coincide with the action. Listen to the resulting speech and effects together with the picture, checking speaker assignment, wording and timing. Synchronized audio is part of the generated clip rather than a separate text answer.\n\n### Size, duration and an image-guided opening\n\nThe standard model’s documented output sizes are 720×1280 portrait and 1280×720 landscape. Higher-resolution sizes in the Pro profile should not be attached to standard Sora 2. Size and duration are explicit request settings; writing a different size or asking for a longer film in the prompt does not change those settings. The creation reference lists 4-, 8- and 12-second clip settings, so choose a short, defined beat when preparing a shot. The expanded producer workflow also documents 16- and 20-second clip settings.\n\nAn image reference anchors the first frame, and the text describes what happens next. The producer supports JPEG, PNG and WebP references and requires the image to match the target video size. Prepare the subject, environment and composition in that opening image, then ask for a specific movement. This is an opening-frame reference rather than a guarantee of matching every detail throughout the clip.\n\n### Reusable non-human characters\n\nThe expanded producer workflow can create a reusable character asset from a short reference video of an animal, mascot or object. Its preparation guidance calls for an MP4 clip around two to four seconds, 720p to 1080p, in 16:9 or 9:16. Matching the source and output aspect ratio helps avoid a stretched appearance. The generation can reference up to two characters.\n\nRetain the character identifier and mention its name in the prompt; passing the identifier alone is not the guide’s complete preparation. Character assets can accompany an image-guided generation, but video extensions do not accept them. The producer blocks human-likeness character uploads by default and rejects reference images containing human faces in the documented API flow. The Sora app’s personal-character features are a separate product workflow.\n\n### Continue or edit a completed clip\n\nAn extension takes a completed source video and a continuation prompt, using the full source clip as context for motion, camera direction and scene continuity. The guide describes additions of up to 20 seconds per extension. Its documented extension input is the source video and text; additional character or image references are excluded from that mode. Describe how the existing action continues instead of introducing an unrelated scene without a transition.\n\nVideo editing targets a clear change to an existing clip. A focused revision can alter an object’s color or the scene’s palette while retaining the rest of the shot. The producer guide recommends one well-defined change at a time and distinguishes editing a generated video from uploading arbitrary footage, which has a separate eligibility condition. Keep the completed source and identify the exact revision needed before making another edit.\n\n### Render jobs and the finished video\n\nThe documented producer video flow is asynchronous: a submitted job can be queued or in progress before it completes or fails. A completed job yields an MP4 video; thumbnail and sprite-sheet assets support previews and browsing. Keep the job identity associated with the prompt, references and returned assets. An initial job response is not the final playable video.\n\nThe ordinary download workflow describes a one-hour download-URL lifetime. The expanded batch workflow prepares independent JSON render requests with stable item identifiers, uploaded references and a result mapped back to each shot; its completed assets have a separate download period of up to 24 hours. Preserve the finished clip promptly and inspect the actual picture and sound before treating it as a selected take.\n\nThe producer’s documented API content restrictions require material suitable for audiences under 18 and reject copyrighted characters or music and generation of real people. Prepare the scene and references accordingly. The direct model result is video with audio, not generated still images, a conversational tool response or an unlimited film. Retain the exact standard-model identity and the accepted source clip when continuing a visual concept.", "editorialOverlay": { "copy_fields": null, "copy_sha256": null, "copy_source": null, "html_sha256": "aa23ec4a1c0a5398eb0761a174e1eb8d3e6ddad72bc4eef51a80c6caf9db2e4d", "module_sha256": "ee9dc1c50b1e6a8fb7fa062030fcb4646b9aaaeb08be513b723a5a92a7290f06", "raw_model_unchanged": true, "schema": "ailoot-native-model-editorial-v1" }, "locale": "en", "methods": [ { "html": "sora-2Sora 2 creates video with synchronized dialogue, sound effects and background audio from text or an image reference. The standard profile supports visual exploration and concept clips, with separate producer workflows for reusable non-human characters, continuation and focused video edits.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Sora 2 is OpenAI’s video-and-audio generation model for dynamic clips created from a written scene description or an image reference. It generates the visual sequence together with synchronized speech, effects and background sound. The standard Sora 2 profile is positioned for visual exploration, concept development, prototypes and rough cuts. Sora 2 Pro is a separate edition aimed at greater visual fidelity and higher-resolution finishing.
The maker’s prompting guide organizes a useful brief around framing, subject, action, setting and light. Give the scene a coherent visual style, then describe the movement in clear beats. An original product concept might show a small mechanical object turning once on a workbench while the camera moves slowly closer, with soft window light and a subtle motor sound. State what the viewer should notice and how the action ends.
Keep each shot’s camera movement and subject action understandable. A complicated sequence can be divided into distinct shot blocks, each with its own framing, action and lighting. Preserve the same subject details and palette when several clips are intended to form one sequence. The guide emphasizes that shorter shots generally follow direction more reliably and that repeated generation of the same prompt produces different takes.
The model’s launch describes improved physical behavior and continuity across shots, including realistic, cinematic and anime styles. These are intended capabilities to direct and inspect. Review motion, contact between objects, proportions and continuity rather than assuming that a plausible first frame guarantees a coherent sequence. If an action is unclear, simplify it and give it a visible beginning and ending.
Specify spoken lines separately from the visual description and label the speakers consistently. Keep dialogue short enough to fit the action and the intended clip length. For a two-character animation, alternate a brief question and answer rather than asking a short shot to accommodate a long speech. The guide warns that complex, lengthy dialogue can disrupt pacing and synchronization.
Describe the soundscape that belongs to the scene: a small movement’s sound, room ambience or the rhythm of an environmental effect. Explain which sound should coincide with the action. Listen to the resulting speech and effects together with the picture, checking speaker assignment, wording and timing. Synchronized audio is part of the generated clip rather than a separate text answer.
The standard model’s documented output sizes are 720×1280 portrait and 1280×720 landscape. Higher-resolution sizes in the Pro profile should not be attached to standard Sora 2. Size and duration are explicit request settings; writing a different size or asking for a longer film in the prompt does not change those settings. The creation reference lists 4-, 8- and 12-second clip settings, so choose a short, defined beat when preparing a shot. The expanded producer workflow also documents 16- and 20-second clip settings.
An image reference anchors the first frame, and the text describes what happens next. The producer supports JPEG, PNG and WebP references and requires the image to match the target video size. Prepare the subject, environment and composition in that opening image, then ask for a specific movement. This is an opening-frame reference rather than a guarantee of matching every detail throughout the clip.
The expanded producer workflow can create a reusable character asset from a short reference video of an animal, mascot or object. Its preparation guidance calls for an MP4 clip around two to four seconds, 720p to 1080p, in 16:9 or 9:16. Matching the source and output aspect ratio helps avoid a stretched appearance. The generation can reference up to two characters.
Retain the character identifier and mention its name in the prompt; passing the identifier alone is not the guide’s complete preparation. Character assets can accompany an image-guided generation, but video extensions do not accept them. The producer blocks human-likeness character uploads by default and rejects reference images containing human faces in the documented API flow. The Sora app’s personal-character features are a separate product workflow.
An extension takes a completed source video and a continuation prompt, using the full source clip as context for motion, camera direction and scene continuity. The guide describes additions of up to 20 seconds per extension. Its documented extension input is the source video and text; additional character or image references are excluded from that mode. Describe how the existing action continues instead of introducing an unrelated scene without a transition.
Video editing targets a clear change to an existing clip. A focused revision can alter an object’s color or the scene’s palette while retaining the rest of the shot. The producer guide recommends one well-defined change at a time and distinguishes editing a generated video from uploading arbitrary footage, which has a separate eligibility condition. Keep the completed source and identify the exact revision needed before making another edit.
The documented producer video flow is asynchronous: a submitted job can be queued or in progress before it completes or fails. A completed job yields an MP4 video; thumbnail and sprite-sheet assets support previews and browsing. Keep the job identity associated with the prompt, references and returned assets. An initial job response is not the final playable video.
The ordinary download workflow describes a one-hour download-URL lifetime. The expanded batch workflow prepares independent JSON render requests with stable item identifiers, uploaded references and a result mapped back to each shot; its completed assets have a separate download period of up to 24 hours. Preserve the finished clip promptly and inspect the actual picture and sound before treating it as a selected take.
The producer’s documented API content restrictions require material suitable for audiences under 18 and reject copyrighted characters or music and generation of real people. Prepare the scene and references accordingly. The direct model result is video with audio, not generated still images, a conversational tool response or an unlimited film. Retain the exact standard-model identity and the accepted source clip when continuing a visual concept.
POST /v1/chat/completionstextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits sora-2 | Default— |
Fieldmessages | Typemessage[] | RequirementRequired | Values & limits | Default— |
Fieldmessages[].role | Typestring | RequirementRequired | Values & limits | Default— |
Fieldmessages[].content | Typestring | RequirementRequired | Values & limits | Default— |
Fieldstream | Typeboolean | RequirementOptional | Values & limits | Defaultfalse |
owned_text_messages
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, messages |
Parametermaterial_types | Contracttext |
Parametermessages | Contractmessage array with role and literal string content |
Parameterfile_material | ContractActual UTF-8 material may be inserted as literal string content; no file reference is inferred. |
Parameteroptional_controls | Contractstream |
Lifecycle: synchronous_json_or_sse
Tariff references do not add request options. Supported parameters are listed for the selected method.
| Configuration | Rate schedule | Charge |
|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Charge Input · Input tokens $0 / 1K tokens Output · Output tokens $0.2 / 1K tokens Input · Cache read $0 / 1K tokens Input · Cache write $0 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model":"sora-2","messages":[{"role":"user","content":"hello"}],"stream":false}')\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/chat/completions \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nfrom pathlib import Path\nprint(Path("result.json").read_text(encoding="utf-8"))\nPY_RESULT\nfrom openai import OpenAI\nimport os\n\nif not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")\nclient=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)\nresult=client.chat.completions.create(model="sora-2", messages=[{"role":"user","content":"hello"}], stream=False)\nimport json\nfrom pathlib import Path\n\ndata = result.to_dict(mode="json")\nPath("result.json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")\nprint(json.dumps(data, ensure_ascii=False, indent=2))\nset -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model":"sora-2","messages":[{"role":"user","content":"hello"}],"stream":true}')\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/chat/completions \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.sse\nJSON or text/event-stream
Keep every complete returned response, file and content part.
Retain the complete JSON or SSE bytes, response headers, returned usage and all supplied result fields. No model-specific output family is asserted.
Preserve the original response and attempt identity; no new job polling or resubmission is inferred.
| Field | Type | Meaning |
|---|---|---|
Fieldcomplete response JSON | Typeobject | MeaningPreserve every returned field without assuming a model-specific result family. |
FieldSSE data events | Typestream | MeaningPreserve every returned event and terminal marker without selecting only text fields. |
Ailoot Ultra sora-2
Use the method paths shown above.
sora-2Sora 2 creates video with synchronized dialogue, sound effects and background audio from text or an image reference. The standard profile supports visual exploration and concept clips, with separate producer workflows for reusable non-human characters, continuation and focused video edits.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Sora 2 is OpenAI’s video-and-audio generation model for dynamic clips created from a written scene description or an image reference. It generates the visual sequence together with synchronized speech, effects and background sound. The standard Sora 2 profile is positioned for visual exploration, concept development, prototypes and rough cuts. Sora 2 Pro is a separate edition aimed at greater visual fidelity and higher-resolution finishing.
The maker’s prompting guide organizes a useful brief around framing, subject, action, setting and light. Give the scene a coherent visual style, then describe the movement in clear beats. An original product concept might show a small mechanical object turning once on a workbench while the camera moves slowly closer, with soft window light and a subtle motor sound. State what the viewer should notice and how the action ends.
Keep each shot’s camera movement and subject action understandable. A complicated sequence can be divided into distinct shot blocks, each with its own framing, action and lighting. Preserve the same subject details and palette when several clips are intended to form one sequence. The guide emphasizes that shorter shots generally follow direction more reliably and that repeated generation of the same prompt produces different takes.
The model’s launch describes improved physical behavior and continuity across shots, including realistic, cinematic and anime styles. These are intended capabilities to direct and inspect. Review motion, contact between objects, proportions and continuity rather than assuming that a plausible first frame guarantees a coherent sequence. If an action is unclear, simplify it and give it a visible beginning and ending.
Specify spoken lines separately from the visual description and label the speakers consistently. Keep dialogue short enough to fit the action and the intended clip length. For a two-character animation, alternate a brief question and answer rather than asking a short shot to accommodate a long speech. The guide warns that complex, lengthy dialogue can disrupt pacing and synchronization.
Describe the soundscape that belongs to the scene: a small movement’s sound, room ambience or the rhythm of an environmental effect. Explain which sound should coincide with the action. Listen to the resulting speech and effects together with the picture, checking speaker assignment, wording and timing. Synchronized audio is part of the generated clip rather than a separate text answer.
The standard model’s documented output sizes are 720×1280 portrait and 1280×720 landscape. Higher-resolution sizes in the Pro profile should not be attached to standard Sora 2. Size and duration are explicit request settings; writing a different size or asking for a longer film in the prompt does not change those settings. The creation reference lists 4-, 8- and 12-second clip settings, so choose a short, defined beat when preparing a shot. The expanded producer workflow also documents 16- and 20-second clip settings.
An image reference anchors the first frame, and the text describes what happens next. The producer supports JPEG, PNG and WebP references and requires the image to match the target video size. Prepare the subject, environment and composition in that opening image, then ask for a specific movement. This is an opening-frame reference rather than a guarantee of matching every detail throughout the clip.
The expanded producer workflow can create a reusable character asset from a short reference video of an animal, mascot or object. Its preparation guidance calls for an MP4 clip around two to four seconds, 720p to 1080p, in 16:9 or 9:16. Matching the source and output aspect ratio helps avoid a stretched appearance. The generation can reference up to two characters.
Retain the character identifier and mention its name in the prompt; passing the identifier alone is not the guide’s complete preparation. Character assets can accompany an image-guided generation, but video extensions do not accept them. The producer blocks human-likeness character uploads by default and rejects reference images containing human faces in the documented API flow. The Sora app’s personal-character features are a separate product workflow.
An extension takes a completed source video and a continuation prompt, using the full source clip as context for motion, camera direction and scene continuity. The guide describes additions of up to 20 seconds per extension. Its documented extension input is the source video and text; additional character or image references are excluded from that mode. Describe how the existing action continues instead of introducing an unrelated scene without a transition.
Video editing targets a clear change to an existing clip. A focused revision can alter an object’s color or the scene’s palette while retaining the rest of the shot. The producer guide recommends one well-defined change at a time and distinguishes editing a generated video from uploading arbitrary footage, which has a separate eligibility condition. Keep the completed source and identify the exact revision needed before making another edit.
The documented producer video flow is asynchronous: a submitted job can be queued or in progress before it completes or fails. A completed job yields an MP4 video; thumbnail and sprite-sheet assets support previews and browsing. Keep the job identity associated with the prompt, references and returned assets. An initial job response is not the final playable video.
The ordinary download workflow describes a one-hour download-URL lifetime. The expanded batch workflow prepares independent JSON render requests with stable item identifiers, uploaded references and a result mapped back to each shot; its completed assets have a separate download period of up to 24 hours. Preserve the finished clip promptly and inspect the actual picture and sound before treating it as a selected take.
The producer’s documented API content restrictions require material suitable for audiences under 18 and reject copyrighted characters or music and generation of real people. Prepare the scene and references accordingly. The direct model result is video with audio, not generated still images, a conversational tool response or an unlimited film. Retain the exact standard-model identity and the accepted source clip when continuing a visual concept.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits sora-2 | Default— |
Fieldmessages | Typemessage[] | RequirementRequired | Values & limits | Default— |
Fieldmessages[].role | Typestring | RequirementRequired | Values & limits user, assistant | Default— |
Fieldmessages[].content | Typestring | RequirementRequired | Values & limits | Default— |
Fieldmax_tokens | Typeinteger | RequirementRequired | Values & limits | Default— |
Fieldstream | Typeboolean | RequirementOptional | Values & limits | Defaultfalse |
Text · UTF-8 text
messages[].content string
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, messages, max_tokens |
Parameteroptional_controls | Contractstream |
Parametermaterial_types | Contracttext |
Parametertext_file_material | ContractRead the actual UTF-8 document and put its text in the indicated input or message content string. |
Lifecycle: synchronous_or_streamed
Tariff references do not add request options. Supported parameters are listed for the selected method.
| Configuration | Rate schedule | Charge |
|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Charge Input · Input tokens $0 / 1K tokens Output · Output tokens $0.2 / 1K tokens Input · Cache read $0 / 1K tokens Input · Cache write $0 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "sora-2",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\n: "${AILOOT_TEXT_FILE:?Set your actual UTF-8 document}"\nexport AILOOT_TEXT_FILE\npython3 - <<'PY_REQUEST'\nimport json, os\nfrom pathlib import Path\n\nsource_file = Path(os.environ['AILOOT_TEXT_FILE'])\nif not source_file.is_file():\n raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')\ntext = source_file.read_text(encoding='utf-8')\nif not text.strip():\n raise SystemExit('Choose a document containing text')\npayload = {'model': 'sora-2', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}\npayload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\\n\\n' + text\nPath('request.json').write_text(json.dumps(payload, ensure_ascii=False), encoding='utf-8')\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/messages \\\n -H "x-api-key: $AILOOT_API_KEY" -H 'anthropic-version: 2023-06-01' -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport json\nfrom pathlib import Path\n\nresult = json.loads(Path('result.json').read_text(encoding='utf-8'))\nprint(json.dumps(result, ensure_ascii=False, indent=2))\nPY_RESULT\nimport json, os\nfrom pathlib import Path\n\nsource_file = Path(os.environ['AILOOT_TEXT_FILE'])\nif not source_file.is_file():\n raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')\ntext = source_file.read_text(encoding='utf-8')\nif not text.strip():\n raise SystemExit('Choose a document containing text')\npayload = {'model': 'sora-2', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}\npayload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\\n\\n' + text\nfrom urllib.request import Request, urlopen\nfrom urllib.error import HTTPError\n\napi_key = os.environ['AILOOT_API_KEY']\nif not api_key.strip():\n raise SystemExit('Set AILOOT_API_KEY before making a request')\nheaders = {'x-api-key': 'PLACEHOLDER_KEY', 'anthropic-version': '2023-06-01', 'Content-Type': 'application/json'}\nheaders['x-api-key'] = api_key\nrequest = Request('https://api.ailoot.ai/v1/messages', data=json.dumps(payload, ensure_ascii=False).encode('utf-8'), headers=headers, method='POST')\ntry:\n with urlopen(request, timeout=120) as response:\n raw = response.read()\n Path('response-headers.txt').write_text(str(response.headers), encoding='utf-8')\nexcept HTTPError as error:\n Path('result.json').write_bytes(error.read())\n Path('response-headers.txt').write_text(str(error.headers), encoding='utf-8')\n raise\nPath('result.json').write_bytes(raw)\nimport json\nfrom pathlib import Path\n\nresult = json.loads(Path('result.json').read_text(encoding='utf-8'))\nprint(json.dumps(result, ensure_ascii=False, indent=2))\nJSON or SSE
Keep every complete returned response, file and content part.
Keep all content blocks, thinking signatures and actual tool-use IDs. Assemble all SSE block deltas; a first text block is not the complete output.
Keep request/response IDs and headers. No generic text job retrieval URL is supplied.
| Field | Type |
|---|---|
Fieldid | Typestring |
Fieldmodel | Typestring |
Fieldcontent[] | Typetext|thinking|tool_use|other content block |
Fieldstop_reason | Typestring|null |
Fieldusage | Typeobject |
| Response | Condition | Next step |
|---|---|---|
Response400 · invalid_request_error | ConditionInvalid required fields, material representation or parameter value | Next stepCorrect the indicated fields before a new creation request. |
Ailoot Ultra sora-2
Use the method paths shown above.