text-embedding-3-large is OpenAI’s larger third-generation text-embedding model for English and non-English tasks. It maps written material to a numerical vector, returning 3,072 components by default. The vector represents relationships in the supplied content; a retrieval or classification process uses it to select material or make another decision.
The third-generation models are trained to support shorter representations through the producer’s dimensions control. The maker gives 1,024 dimensions as an example when a vector store requires that size. A shorter vector reduces the amount of data to store and compare, while changing the representation can affect retrieval quality. Choose the size for the application’s requirements and use that same choice for the corpus and its queries.
An illustrative document-search workflow divides source material into meaningful passages, retains their identifiers and obtains vectors for each passage. A question is embedded using the same model and dimensions, then compared with the passage vectors. The application supplies filters, selects the ranked results and shows their source references. A subsequent text-generation model can develop an answer from the retrieved material; the embedding model’s result remains numerical data.
Other documented applications include code or text search, grouping similar records, related-content recommendations, classification, anomaly detection and measuring diversity in a collection. For a recommendation, describe the objects in the terms relevant to the relationship. For classification, compare against category representations or use vectors as features in a classifier. Define the rule that turns similarity into the result needed by the product.
OpenAI lists an 8,192-token maximum input per text. The embedding API accepts strings or tokenized text, supports multiple inputs and rejects empty strings. It also enforces a 300,000-token aggregate limit for one request. A long document should be divided with its meaningful boundaries and context intact, keeping each vector connected to its source passage.
The producer documents float vectors as the default output and base64 as an alternative encoding. The guide recommends cosine similarity and states that OpenAI embeddings are normalized to unit length, allowing a dot product for cosine similarity and equivalent rankings under Euclidean distance. If a third-generation vector is shortened manually after generation, the maker instructs the application to normalize the shortened vector again; using the dimensions control is the recommended route.
Prepare a consistent text representation and record the model, dimensions and source identifier. A useful final artifact is a ranked set of passages, related items or grouped records whose selection rule is explicit. Assess whether the result addresses the query and preserve the source evidence for the next stage, rather than treating vector similarity as verification of a record’s factual claims.
", "aboutText": "text-embedding-3-large is OpenAI’s larger third-generation text-embedding model for English and non-English tasks. It maps written material to a numerical vector, returning 3,072 components by default. The vector represents relationships in the supplied content; a retrieval or classification process uses it to select material or make another decision.\n\nThe third-generation models are trained to support shorter representations through the producer’s dimensions control. The maker gives 1,024 dimensions as an example when a vector store requires that size. A shorter vector reduces the amount of data to store and compare, while changing the representation can affect retrieval quality. Choose the size for the application’s requirements and use that same choice for the corpus and its queries.\n\nAn illustrative document-search workflow divides source material into meaningful passages, retains their identifiers and obtains vectors for each passage. A question is embedded using the same model and dimensions, then compared with the passage vectors. The application supplies filters, selects the ranked results and shows their source references. A subsequent text-generation model can develop an answer from the retrieved material; the embedding model’s result remains numerical data.\n\nOther documented applications include code or text search, grouping similar records, related-content recommendations, classification, anomaly detection and measuring diversity in a collection. For a recommendation, describe the objects in the terms relevant to the relationship. For classification, compare against category representations or use vectors as features in a classifier. Define the rule that turns similarity into the result needed by the product.\n\nOpenAI lists an 8,192-token maximum input per text. The embedding API accepts strings or tokenized text, supports multiple inputs and rejects empty strings. It also enforces a 300,000-token aggregate limit for one request. A long document should be divided with its meaningful boundaries and context intact, keeping each vector connected to its source passage.\n\nThe producer documents float vectors as the default output and base64 as an alternative encoding. The guide recommends cosine similarity and states that OpenAI embeddings are normalized to unit length, allowing a dot product for cosine similarity and equivalent rankings under Euclidean distance. If a third-generation vector is shortened manually after generation, the maker instructs the application to normalize the shortened vector again; using the dimensions control is the recommended route.\n\nPrepare a consistent text representation and record the model, dimensions and source identifier. A useful final artifact is a ranked set of passages, related items or grouped records whose selection rule is explicit. Assess whether the result addresses the query and preserve the source evidence for the next stage, rather than treating vector similarity as verification of a record’s factual claims.", "editorialOverlay": { "copy_fields": null, "copy_sha256": null, "copy_source": null, "html_sha256": "7a481df47d802c16657309a9a80a1ea5b5f66040e106d32b6b7e019f5477f585", "module_sha256": "ee9dc1c50b1e6a8fb7fa062030fcb4646b9aaaeb08be513b723a5a92a7290f06", "raw_model_unchanged": true, "schema": "ailoot-native-model-editorial-v1" }, "locale": "en", "methods": [ { "html": "openai/text-embedding-3-largeOpenAI’s larger third-generation text-embedding model returns 3,072-dimensional vectors by default, with a producer control for shorter representations. Use it for semantic retrieval, multilingual comparison and related-content or classification workflows.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
text-embedding-3-large is OpenAI’s larger third-generation text-embedding model for English and non-English tasks. It maps written material to a numerical vector, returning 3,072 components by default. The vector represents relationships in the supplied content; a retrieval or classification process uses it to select material or make another decision.
The third-generation models are trained to support shorter representations through the producer’s dimensions control. The maker gives 1,024 dimensions as an example when a vector store requires that size. A shorter vector reduces the amount of data to store and compare, while changing the representation can affect retrieval quality. Choose the size for the application’s requirements and use that same choice for the corpus and its queries.
An illustrative document-search workflow divides source material into meaningful passages, retains their identifiers and obtains vectors for each passage. A question is embedded using the same model and dimensions, then compared with the passage vectors. The application supplies filters, selects the ranked results and shows their source references. A subsequent text-generation model can develop an answer from the retrieved material; the embedding model’s result remains numerical data.
Other documented applications include code or text search, grouping similar records, related-content recommendations, classification, anomaly detection and measuring diversity in a collection. For a recommendation, describe the objects in the terms relevant to the relationship. For classification, compare against category representations or use vectors as features in a classifier. Define the rule that turns similarity into the result needed by the product.
OpenAI lists an 8,192-token maximum input per text. The embedding API accepts strings or tokenized text, supports multiple inputs and rejects empty strings. It also enforces a 300,000-token aggregate limit for one request. A long document should be divided with its meaningful boundaries and context intact, keeping each vector connected to its source passage.
The producer documents float vectors as the default output and base64 as an alternative encoding. The guide recommends cosine similarity and states that OpenAI embeddings are normalized to unit length, allowing a dot product for cosine similarity and equivalent rankings under Euclidean distance. If a third-generation vector is shortened manually after generation, the maker instructs the application to normalize the shortened vector again; using the dimensions control is the recommended route.
Prepare a consistent text representation and record the model, dimensions and source identifier. A useful final artifact is a ranked set of passages, related items or grouped records whose selection rule is explicit. Assess whether the result addresses the query and preserve the source evidence for the next stage, rather than treating vector similarity as verification of a record’s factual claims.
POST /v1/embeddingstext, token_idsembedding| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits openai/text-embedding-3-large | Default— |
Fieldinput | Typestring|string[]|integer[]|integer[][] | RequirementRequired | Values & limits Minimum items: 1 Conditions: Nonempty homogeneous batch. Token IDs are nonnegative integers. Input budget: Maximum: 8,192 · Text unit: UTF8 bytes · Token unit: token-array elements | Default— |
Fieldencoding_format | Typestring | RequirementOptional | Values & limits float | Defaultfloat |
Fielddimensions | Typeinteger | RequirementNot supported | Values & limits Minimum: 1 | Default— |
Fielduser | Typejson_value | RequirementOptional | Values & limits Conditions: Source allows this JSON metadata field; exact upstream semantics are model-specific. | Default— |
Fieldsession_id | Typejson_value | RequirementOptional | Values & limits Conditions: Source allows this JSON metadata field; exact upstream semantics are model-specific. | Default— |
Fieldtrace | Typejson_value | RequirementOptional | Values & limits Conditions: Source allows this JSON metadata field; exact upstream semantics are model-specific. | Default— |
Fieldprovider | Typeobject | RequirementOptional | Values & limits Conditions: provider.options rejected | Default— |
text_to_embedding
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, input |
Parameterinput_shapes | Contractstring, string[], nonnegative_integer_token_array, nonnegative_integer_token_array[] |
Parameterencoding_format | Contractfloat |
Parameterdimensions_override | Contractfalse |
Parametermax_input_budget_units | Contract8192 |
Parametermultimodal_objects | Contractfalse |
Parametermaterial_types | Contracttext, token_ids |
Lifecycle: synchronous_json
Keep every vector index and dimension.
Multimodal model name does not admit image objects on this exact path.
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 $0.00026 / 1K tokens Output · Output tokens $0 / 1K tokens Input · Cache read — / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model": "openai/text-embedding-3-large", "input": ["First document", "Second document"]}')\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/embeddings \\\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 json\nfrom pathlib import Path\nr=json.loads(Path("result.json").read_text())\nfor x in r.get("data",[]): print(json.dumps({"index":x.get("index"),"embedding":x.get("embedding")}))\nPY_RESULT\nfrom openai import OpenAI\nimport os\n\nclient=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)\nresult=client.embeddings.create(model='openai/text-embedding-3-large',input=["First document","Second document"])\nprint(result)\nfor item in result.data:print(item.index,item.embedding)\njson
Keep every complete returned response, file and content part.
Keep every vector with its index. Vector count must equal input count; dimensions must be consistent.
No embedding job endpoint is introduced.
| Field | Type |
|---|---|
Fieldobject | Typestring |
Fieldmodel | Typestring |
Fielddata[].index | Typeinteger |
Fielddata[].embedding | Typefinite float[] |
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 openai/text-embedding-3-large, text-embedding-3-large
Use the method paths shown above.
text-embedding-3-largeOpenAI’s larger third-generation text-embedding model returns 3,072-dimensional vectors by default, with a producer control for shorter representations. Use it for semantic retrieval, multilingual comparison and related-content or classification workflows.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
text-embedding-3-large is OpenAI’s larger third-generation text-embedding model for English and non-English tasks. It maps written material to a numerical vector, returning 3,072 components by default. The vector represents relationships in the supplied content; a retrieval or classification process uses it to select material or make another decision.
The third-generation models are trained to support shorter representations through the producer’s dimensions control. The maker gives 1,024 dimensions as an example when a vector store requires that size. A shorter vector reduces the amount of data to store and compare, while changing the representation can affect retrieval quality. Choose the size for the application’s requirements and use that same choice for the corpus and its queries.
An illustrative document-search workflow divides source material into meaningful passages, retains their identifiers and obtains vectors for each passage. A question is embedded using the same model and dimensions, then compared with the passage vectors. The application supplies filters, selects the ranked results and shows their source references. A subsequent text-generation model can develop an answer from the retrieved material; the embedding model’s result remains numerical data.
Other documented applications include code or text search, grouping similar records, related-content recommendations, classification, anomaly detection and measuring diversity in a collection. For a recommendation, describe the objects in the terms relevant to the relationship. For classification, compare against category representations or use vectors as features in a classifier. Define the rule that turns similarity into the result needed by the product.
OpenAI lists an 8,192-token maximum input per text. The embedding API accepts strings or tokenized text, supports multiple inputs and rejects empty strings. It also enforces a 300,000-token aggregate limit for one request. A long document should be divided with its meaningful boundaries and context intact, keeping each vector connected to its source passage.
The producer documents float vectors as the default output and base64 as an alternative encoding. The guide recommends cosine similarity and states that OpenAI embeddings are normalized to unit length, allowing a dot product for cosine similarity and equivalent rankings under Euclidean distance. If a third-generation vector is shortened manually after generation, the maker instructs the application to normalize the shortened vector again; using the dimensions control is the recommended route.
Prepare a consistent text representation and record the model, dimensions and source identifier. A useful final artifact is a ranked set of passages, related items or grouped records whose selection rule is explicit. Assess whether the result addresses the query and preserve the source evidence for the next stage, rather than treating vector similarity as verification of a record’s factual claims.
POST /v1/embeddingstextembedding| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits text-embedding-3-large | Default— |
Fieldinput | Typestring|string[]|integer[]|integer[][] | RequirementRequired | Values & limits Minimum items: 1 Conditions: Nonempty homogeneous batch. Token IDs are nonnegative integers. Input budget: Maximum: 8,192 · Text unit: UTF8 bytes · Token unit: token-array elements | Default— |
Fieldencoding_format | Typestring | RequirementOptional | Values & limits float | Defaultfloat |
Fielddimensions | Typeinteger | RequirementNot supported | Values & limits Minimum: 1 | Default— |
Fielduser | Typejson_value | RequirementOptional | Values & limits Conditions: Optional JSON metadata; exact upstream semantics depend on the selected model. | Default— |
Fieldsession_id | Typejson_value | RequirementOptional | Values & limits Conditions: Optional JSON metadata; exact upstream semantics depend on the selected model. | Default— |
Fieldtrace | Typejson_value | RequirementOptional | Values & limits Conditions: Optional JSON metadata; exact upstream semantics depend on the selected model. | Default— |
Fieldprovider | Typeobject | RequirementOptional | Values & limits Conditions: provider.options rejected | Default— |
text_to_embedding
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, input |
Parameterinput_shapes | Contractstring, string[], nonnegative_integer_token_array, nonnegative_integer_token_array[] |
Parameterencoding_format | Contractfloat |
Parameterdimensions_override | Contractfalse |
Parametermax_input_budget_units | Contract8192 |
Parametermultimodal_objects | Contractfalse |
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_json
Send the exact model ID and required fields.
Keep every returned item and its original identifiers.
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 $0.00026 / 1K tokens Output · Output tokens $0 / 1K tokens Input · Cache read — / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "text-embedding-3-large",\n "input": [\n "First document",\n "Second document"\n ],\n "encoding_format": "float"\n}set -eu\n: "${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': 'text-embedding-3-large', 'input': ['First document', 'Second document'], 'encoding_format': 'float'}\npayload['input'] = [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/embeddings \\\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 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))\nfor vector in result.get('data', []):\n print('Vector', vector.get('index'), json.dumps(vector.get('embedding')))\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': 'text-embedding-3-large', 'input': ['First document', 'Second document'], 'encoding_format': 'float'}\npayload['input'] = [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 = {'Authorization': 'Bearer PLACEHOLDER_KEY', 'Content-Type': 'application/json'}\nheaders['Authorization'] = 'Bearer ' + api_key\nrequest = Request('https://api.ailoot.ai/v1/embeddings', 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))\nfor vector in result.get('data', []):\n print('Vector', vector.get('index'), json.dumps(vector.get('embedding')))\njson
Keep every complete returned response, file and content part.
Keep every finite vector with its original index. Preserve all dimensions and the complete usage object.
No embedding job endpoint is introduced.
| Field | Type |
|---|---|
Fieldobject | Typestring |
Fieldmodel | Typestring |
Fielddata[].index | Typeinteger |
Fielddata[].embedding | Typefinite float[] |
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 openai/text-embedding-3-large, text-embedding-3-large
Use the method paths shown above.