Qwen3 Reranker 8B is a text reranking model available under the exact ID qwen/qwen3-reranker-8b. It takes a query and a collection of candidate documents and returns a relevance ranking. A practical place for it is between an initial search and the application that consumes the selected passages: retrieve possible matches first, rerank those matches against the actual question, and pass the chosen source text to the next step.
Send POST /v1/rerank with a JSON body containing model, query and documents. Set model to qwen/qwen3-reranker-8b, supply the search question as a string, and put the candidate passages in a nonempty array of text strings. Preserve the input array order and your own document identifiers alongside it, because the returned index identifies a document by its original position. Include the material that determines relevance, such as a product name, policy condition or technical passage, in the text supplied for that candidate.
The optional top_n parameter is an integer with a minimum of 1 and selects how many of the most relevant documents to return. For a support search, the query could be a customer’s exact question and the documents could be candidate help-centre sections. For an internal knowledge base, use the question together with passages that retain their section titles and enough surrounding text to explain each result. Choose a result count that fits the downstream answer or search display.
The published model profile lists a context length of 40,960 tokens. Account for the query and supplied document text when preparing the request, and select the passages that are useful for the question. Reranking produces scores and source selections; use the selected document text when another step needs to compose an answer or quotation.
The operation returns a complete JSON response. Read results in their returned relevance order, retain each result’s index, relevance_score and returned document content, and keep the response model, identifier and usage fields when present. Map indices back to your original documents before presenting a source or assembling an answer. Save the original response and request identity if delivery is interrupted so the existing result can be investigated without submitting the same uncertain operation again.
", "aboutText": "Qwen3 Reranker 8B is a text reranking model available under the exact ID qwen/qwen3-reranker-8b. It takes a query and a collection of candidate documents and returns a relevance ranking. A practical place for it is between an initial search and the application that consumes the selected passages: retrieve possible matches first, rerank those matches against the actual question, and pass the chosen source text to the next step.\n\nSend POST /v1/rerank with a JSON body containing model, query and documents. Set model to qwen/qwen3-reranker-8b, supply the search question as a string, and put the candidate passages in a nonempty array of text strings. Preserve the input array order and your own document identifiers alongside it, because the returned index identifies a document by its original position. Include the material that determines relevance, such as a product name, policy condition or technical passage, in the text supplied for that candidate.\n\nThe optional top_n parameter is an integer with a minimum of 1 and selects how many of the most relevant documents to return. For a support search, the query could be a customer’s exact question and the documents could be candidate help-centre sections. For an internal knowledge base, use the question together with passages that retain their section titles and enough surrounding text to explain each result. Choose a result count that fits the downstream answer or search display.\n\nThe published model profile lists a context length of 40,960 tokens. Account for the query and supplied document text when preparing the request, and select the passages that are useful for the question. Reranking produces scores and source selections; use the selected document text when another step needs to compose an answer or quotation.\n\nThe operation returns a complete JSON response. Read results in their returned relevance order, retain each result’s index, relevance_score and returned document content, and keep the response model, identifier and usage fields when present. Map indices back to your original documents before presenting a source or assembling an answer. Save the original response and request identity if delivery is interrupted so the existing result can be investigated without submitting the same uncertain operation again.", "editorialOverlay": { "copy_fields": null, "copy_sha256": null, "copy_source": null, "html_sha256": "ca9216b223c8d569fe61bc225e0b1a36c3f3a96a4002a59c1ea10f59838417dd", "module_sha256": "ee9dc1c50b1e6a8fb7fa062030fcb4646b9aaaeb08be513b723a5a92a7290f06", "raw_model_unchanged": true, "schema": "ailoot-native-model-editorial-v1" }, "locale": "en", "methods": [ { "html": "qwen/qwen3-reranker-8bQwen3 Reranker 8B ranks supplied text documents against a search query. Send qwen/qwen3-reranker-8b to the Rerank endpoint with the query and candidate passages, then use the returned relevance scores and original document indices to select material for search results, a knowledge-base answer or a retrieval workflow.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Qwen3 Reranker 8B is a text reranking model available under the exact ID qwen/qwen3-reranker-8b. It takes a query and a collection of candidate documents and returns a relevance ranking. A practical place for it is between an initial search and the application that consumes the selected passages: retrieve possible matches first, rerank those matches against the actual question, and pass the chosen source text to the next step.
Send POST /v1/rerank with a JSON body containing model, query and documents. Set model to qwen/qwen3-reranker-8b, supply the search question as a string, and put the candidate passages in a nonempty array of text strings. Preserve the input array order and your own document identifiers alongside it, because the returned index identifies a document by its original position. Include the material that determines relevance, such as a product name, policy condition or technical passage, in the text supplied for that candidate.
The optional top_n parameter is an integer with a minimum of 1 and selects how many of the most relevant documents to return. For a support search, the query could be a customer’s exact question and the documents could be candidate help-centre sections. For an internal knowledge base, use the question together with passages that retain their section titles and enough surrounding text to explain each result. Choose a result count that fits the downstream answer or search display.
The published model profile lists a context length of 40,960 tokens. Account for the query and supplied document text when preparing the request, and select the passages that are useful for the question. Reranking produces scores and source selections; use the selected document text when another step needs to compose an answer or quotation.
The operation returns a complete JSON response. Read results in their returned relevance order, retain each result’s index, relevance_score and returned document content, and keep the response model, identifier and usage fields when present. Map indices back to your original documents before presenting a source or assembling an answer. Save the original response and request identity if delivery is interrupted so the existing result can be investigated without submitting the same uncertain operation again.
POST /v1/reranktextranking| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits qwen/qwen3-reranker-8b | Default— |
Fieldquery | Typestring | RequirementRequired | Values & limits | Default— |
Fielddocuments | Typearray | RequirementRequired | Values & limits | Default— |
Fieldtop_n | Typeinteger | RequirementOptional | Values & limits Minimum: 1 | Default— |
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, query, documents |
Parameteroptional | Contracttop_n |
Parametermaterial_types | Contracttext |
Parameterconditions | Contract |
Lifecycle: synchronous_json
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 | Billing unit | Charge |
|---|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Billing unit— | Charge Input · Input tokens — / 1K tokens Output · Output tokens — / 1K tokens Input · Cache read — / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
| Configurationinput token | Rate schedule— | Billing unit/ 1K tokens | Charge Input · input token $0.0004 / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\nexport AILOOT_API_KEY\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"documents": ["Paris is the capital of France.", "Berlin is the capital of Germany."], "model": "qwen/qwen3-reranker-8b", "query": "What is the capital of France?", "top_n": 3}')\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/rerank \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\n# Submit once. Retain the complete result, original identifiers and every response header.\nJSON
Keep every complete returned response, file and content part.
Retain the complete response, all actual results, original identifiers, usage and response headers.
Preserve the original response, headers and X-Client-Request-ID. An uncertain outcome does not authorize another creation.
| Field | Type |
|---|---|
Fieldresults | Typearray |
Ailoot Ultra qwen/qwen3-reranker-8b
Use the method paths shown above.