# Duplicate index
- Operation ID: `duplicateIndexV2Get`
- HTTP method: `GET`
- Path: `/search/v2/indices/{indexId}/duplicate`
- [Human-readable API reference](https://hub.synerise.com/api-reference/ai-search#tag/Search-Configuration/operation/duplicateIndexV2Get)
## Self-contained OpenAPI method
The fenced document below contains this method's documentation and all of its local references. It is self-contained; no category or master specification fetch is required.
```yaml
openapi: 3.0.0
info:
title: Synerise Public API
version: 1.9.1
paths:
/search/v2/indices/{indexId}/duplicate:
get:
summary: Duplicate index
description: |
duplicate index.
---
**API consumers:** Workspace (Business Profile), Synerise User
**API key permission required:** `ITEMS_SEARCH_CONFIG_SEARCH_CREATE`
**User role permission required:** `assets_search: create`
operationId: duplicateIndexV2Get
tags:
- Search Configuration
security:
- JWT: []
parameters:
- in: path
name: indexId
required: true
description: ID of the index
schema:
type: string
responses:
"200":
description: Returned newly created duplicate index
content:
application/json:
schema:
type: object
description: Details of a single index
properties:
indexId:
type: string
description: ID of the index
indexName:
type: string
description: Human-friendly name of the index
description:
type: string
description: Description of the index
itemsCatalogId:
type: string
description: ID of the item catalog from which the index will be created
language:
type: string
description: Search language as ISO 639-1 code
enabled:
type: boolean
description: When `true`, the index is enabled and can be queried.
ignoreUnavailableItems:
type: boolean
description: When `true`, unavailable items are not indexed, which makes the search run faster.
default: true
scoring:
description: |
Item scoring settings that affect the presentation order of the results.
`0` means the lowest importance of a parameter, `1` is the highest importance.
type: object
properties:
pageVisitsPopularity:
description: The importance of how often the item page is visited
type: number
format: float
default: 0
minimum: 0
maximum: 1
transactionsPopularity:
description: The importance of how often the item is included in transactions
type: number
format: float
default: 0
minimum: 0
maximum: 1
personalized:
description: The importance of how relevant a item is to a profile according to the AI model
type: number
format: float
default: 0
minimum: 0
maximum: 1
tieBreaker:
description: The importance of lesser phrase matches
type: number
format: float
default: 0
minimum: 0
maximum: 1
matching:
description: Controls how many clauses should be present in query
oneOf:
- type: object
description: Indicates number of mandatory clauses
properties:
type:
type: string
enum:
- MinimumMatching
value:
type: integer
minimum: 1
- type: object
description: Indicates number of optional clauses
properties:
type:
type: string
enum:
- MaximumNotMatching
value:
type: integer
minimum: 1
- type: object
description: Any of clauses should match
properties:
type:
type: string
enum:
- AnyMatching
- type: object
description: All clauses should match
properties:
type:
type: string
enum:
- AllMatching
- type: object
description: |
Matching that works differently depending on the number of terms in the search query.
properties:
type:
type: string
enum:
- AdaptiveMatching
termsToMatch:
type: number
format: float
minimum: 0
exclusiveMinimum: true
maximum: 1
exclusiveMaximum: true
description: |
Indicates the percentage of all query terms that should be matched.
This matching is done using MinimumMatching with the value set to:
`termsToMatch * n`, where n is a number of all terms in the search query.
**If you define termsToMatch, you can't define termsToSubtract.**
termsToSubtract:
type: integer
minimum: 1
default: 1
description: |
Indicates a fixed number that should be subtracted from
the number of query terms that should match. This matching is done using
MinimumMatching with the value set to `n - termsToSubtract`, where n is
a number of all terms in the search query.
**If you define termsToSubtract, you can't define termsToMatch.**
termsToActivate:
type: integer
minimum: 1
default: 3
description: |
Indicates how long (number of words) a search query must be to activate this mechanism.
All queries that are shorter than termsToActivate will use AllMatching.
fallbackMatching:
description: |
Backup matching mode used in cases where the main matching mode
(specified in `matching` parameter) does not provide enough results (at least 5)
oneOf:
- type: object
description: Indicates number of mandatory clauses
properties:
type:
type: string
enum:
- MinimumMatching
value:
type: integer
minimum: 1
- type: object
description: Indicates number of optional clauses
properties:
type:
type: string
enum:
- MaximumNotMatching
value:
type: integer
minimum: 1
- type: object
description: Any of clauses should match
properties:
type:
type: string
enum:
- AnyMatching
- type: object
description: |
Matching that works differently depending on the number of terms in the search query.
properties:
type:
type: string
enum:
- AdaptiveMatching
termsToMatch:
type: number
format: float
minimum: 0
exclusiveMinimum: true
maximum: 1
exclusiveMaximum: true
description: |
Indicates the percentage of all query terms that should be matched.
This matching is done using MinimumMatching with the value set to:
`termsToMatch * n`, where n is a number of all terms in the search query.
**If you define termsToMatch, you can't define termsToSubtract.**
termsToSubtract:
type: integer
minimum: 1
default: 1
description: |
Indicates a fixed number that should be subtracted from
the number of query terms that should match. This matching is done using
MinimumMatching with the value set to `n - termsToSubtract`, where n is
a number of all terms in the search query.
**If you define termsToSubtract, you can't define termsToMatch.**
termsToActivate:
type: integer
minimum: 1
default: 3
description: |
Indicates how long (number of words) a search query must be to activate this mechanism.
All queries that are shorter than termsToActivate will use AllMatching.
similarity:
description: Defines how matching documents are scored
oneOf:
- type: object
description: Enhanced TF-IDF scoring model
properties:
type:
type: string
enum:
- BM25
k1:
type: number
description: Controls non-linear term frequency normalization (saturation)
format: float
default: 1.2
minimum: 0
maximum: 10
b:
type: number
description: Controls to what degree document length normalizes tf values
format: float
default: 0.75
minimum: 0
maximum: 1
- type: object
description: Term frequency-inverse document frequency scoring model
properties:
type:
type: string
enum:
- IDF
crossWorkspaceMode:
description: Cross workspace mode configuration
type: object
required:
- enabled
properties:
enabled:
description: When `true`, cross workspace search personalization mode is enabled
type: boolean
suggestions:
type: object
description: Controls manner in which suggestions are generated
properties:
smoothingModel:
type: string
description: Controls balance weight between infrequent grams and frequent grams
enum:
- Popularity
- Accuracy
gramSize:
type: integer
description: Describes max size of the n-grams in the field
default: 3
minimum: 1
maximum: 10
maxErrors:
type: integer
description: Maximum number of errors for terms to be considered as misspelling in order to for a correction
default: 2
minimum: 1
maximum: 5
minWordLength:
type: integer
description: The minimum length a suggest text term must have in order to be included
default: 4
minimum: 1
useAlways:
type: boolean
description: Controls is suggestions are used only if no documents match, or always
default: false
tokenizer:
description: Used for tokenizing full text into individual words
oneOf:
- type: object
description: The whitespace tokenizer divides text into terms whenever it encounters any whitespace character
properties:
type:
type: string
enum:
- Whitespace
- type: object
description: The standard tokenizer divides text into terms on word boundaries
properties:
type:
type: string
enum:
- Standard
- type: object
description: The pattern tokenizer uses a regular expression to either split text into terms whenever it matches a word separator
properties:
type:
type: string
enum:
- Pattern
pattern:
type: string
example: "[.,;]"
analyzers:
type: object
description: Controls analyzer settings.
required:
- autocomplete
properties:
autocomplete:
type: string
description: The completion method used.
enum:
- Ngram
- EdgeNgram
- NgramWithSynonyms
- EdgeNgramWithSynonyms
- DefaultWithSynonyms
attributesWithoutPrefixSearch:
type: array
description: Searchable attributes which will not be used in a prefix search
example:
- brand
- color
- promotion.type
items:
type: string
attributesWithoutTypoTolerance:
type: array
description: Searchable attributes for which typo tolerance is off
example:
- itemId
- link
items:
type: string
valuesWithoutTypoTolerance:
type: array
description: Attributes values for which typo tolerance is off
example:
- Nike
- blue
- "74997"
items:
type: string
typoToleranceOnNumericValues:
type: boolean
description: When `true`, typo tolerance is active on numbers
default: true
searchableAttributes:
description: |
The attributes that are taken into account as full text search terms
type: object
properties:
fullText:
type: array
description: The least important parameters
example:
- brand
- color
- promotion.type
items:
type: string
fullTextBoosted:
type: array
description: Medium-importance parameters
example:
- brand
- color
- promotion.type
items:
type: string
fullTextStronglyBoosted:
type: array
description: The most important parameters
example:
- brand
- color
- promotion.type
items:
type: string
displayableAttributes:
description: Attributes shown in the search results
example:
- brand
- color
- promotion.type
type: array
items:
type: string
facetableAttributes:
description: Attributes for which facets are returned in the search response. By default, all facetable attributes are also filterable and sortable.
type: object
properties:
range:
type: array
description: Array of range attributes
example:
- age
- customScore.value
- salePrice.value
default: []
items:
type: string
text:
type: array
description: Array of text attributes
example:
- brand
- color
- promotion.type
default: []
items:
type: string
datetime:
type: array
description: Array of datetime attributes
example:
- created
- attributes.modifiedAt
default: []
items:
type: string
filterableAttributes:
description: Attributes for which filters can be used
type: object
properties:
range:
type: array
description: Array of range attributes
example:
- age
- customScore.value
- salePrice.value
default: []
items:
type: string
text:
type: array
description: Array of text attributes
example:
- brand
- color
- promotion.type
default: []
items:
type: string
datetime:
type: array
description: Array of datetime attributes
example:
- created
- attributes.modifiedAt
default: []
items:
type: string
sortableAttributes:
description: Attributes using which search results can be sorted
type: object
properties:
range:
type: array
description: Array of range attributes
example:
- age
- customScore.value
- salePrice.value
default: []
items:
type: string
text:
type: array
description: Array of text attributes
example:
- brand
- color
- promotion.type
default: []
items:
type: string
datetime:
type: array
description: Array of datetime attributes
example:
- created
- attributes.modifiedAt
default: []
items:
type: string
distinctFilterAttributes:
description: Attributes for which distinct can be used
type: object
properties:
range:
type: array
description: Array of range attributes
example:
- age
- customScore.value
- salePrice.value
default: []
items:
type: string
text:
type: array
description: Array of text attributes
example:
- brand
- color
- promotion.type
default: []
items:
type: string
datetime:
type: array
description: Array of datetime attributes
example:
- created
- attributes.modifiedAt
default: []
items:
type: string
recentSearches:
type: object
description: Recent searches configuration
properties:
windowSize:
description: Amount of recent searches items to be returned
type: integer
example: 100
timeSpan:
type: object
description: Defines period for which recent searches are to be returned
properties:
type:
description: One of the following time units
type: string
enum:
- YEARS
- MONTHS
- WEEKS
- DAYS
- HOURS
- MINUTES
- SECONDS
example: DAYS
value:
description: Value corresponding with the time unit
type: integer
example: 30
reranking:
type: object
description: Reranking configuration
properties:
dynamicReranker:
type: object
description: Dynamic reranker configuration
required:
- enabled
- numberOfItems
properties:
enabled:
type: boolean
description: When `true`, the dynamic reranker is enabled
filter:
type: string
description: Only items that meet this filter will be considered by the reranker
numberOfItems:
type: integer
description: Number of items that can be reranked in search results
minimum: 1
maximum: 50
minNumberOfEvents:
type: integer
description: If present, items with fewer events than this value will be filtered out from reranking.
queryClassification:
type: object
description: Query classification configuration
properties:
enabled:
type: boolean
description: When `true`, the query classification is enabled.
actions:
type: array
description: Search result actions with a minimum reliability level
items:
type: object
properties:
action:
type: string
description: |
The action to perform when the reliability level from `minReliabilityLevel` is reached.
- `staticFilter` will apply a static filter to limit the search results to the predicted category.
- `elasticFilter` will apply an elastic filter - if there are not enough results from the predicted category, the AI engine will pick items from other categories to meet the minimum number of results.
- `boosting` will increase the scores of items from the predicted category.
If reliability levels are met for multiple actions, the priority is:
1. static filter
2. elastic filter
3. boosting
enum:
- staticFilter
- elasticFilter
- boosting
minReliabilityLevel:
type: string
description: |
For each type of action, you can specify the minimum level required to trigger that action. A reliability level indicates how confident we are that a given prediction is the best choice.
enum:
- certain
- veryHigh
- high
- medium
- low
excludedPhrases:
type: array
description: Phrases that will be excluded from query classification
items:
type: string
excludedCategories:
type: array
description: Categories that will be excluded from query classification
items:
type: string
optimization:
type: object
description: Configuration for the search weight optimizer
properties:
enabled:
type: boolean
description: Gate for the optimization feature. items-search applies `optimizedWeights` (via `target` or a query-rule target) only when `true`.
target:
type: string
description: |
The optimization target for the weight optimizer. Specifies which metric to optimize for.
- `ctr` — Click-through rate
- `cr` — Conversion rate
- `revenue` — Revenue
- `ndcg` — Normalized Discounted Cumulative Gain
- `margin` — Margin
enum:
- ctr
- cr
- revenue
- ndcg
- margin
optimizedWeights:
type: object
description: Per-target optimization weight configuration, keyed by optimization target (ctr, cr, revenue, ndcg, margin)
additionalProperties:
type: object
description: Optimization weights for a specific target, containing optional base and DRR weights
properties:
baseWeights:
type: object
description: Base weight configuration for optimization
properties:
personalized:
type: number
format: float
description: Weight for personalization factor
minimum: 0
maximum: 5
transactionsPopularity:
type: number
format: float
description: Weight for transactions popularity
minimum: 0
maximum: 5
pageVisitsPopularity:
type: number
format: float
description: Weight for page visits popularity
minimum: 0
maximum: 5
transactions7DaysPopularity:
type: number
format: float
description: Weight for transactions popularity in the last 7 days
minimum: 0
maximum: 5
pageVisits7DaysPopularity:
type: number
format: float
description: Weight for page visits popularity in the last 7 days
minimum: 0
maximum: 5
revenueBoost:
type: number
format: float
description: Weight for revenue boost factor
minimum: 0
maximum: 5
conversionRateBoost:
type: number
format: float
description: Weight for conversion rate boost factor
minimum: 0
maximum: 5
fullTextMultipliers:
type: object
description: |
Per-field-tier multipliers applied on top of the default full-text field boosts. Each tier is optional; an absent tier means no change (multiplier 1.0). Scale a tier down to stop it dominating (e.g. a rare description token outscoring a title match) or up to make it decisive.
properties:
stronglyBoosted:
type: number
format: float
description: Multiplier for the strongly-boosted tier (title, brand)
minimum: 0
maximum: 100
boosted:
type: number
format: float
description: Multiplier for the boosted tier (category)
minimum: 0
maximum: 100
normal:
type: number
format: float
description: Multiplier for the normal tier (description, color, itemId)
minimum: 0
maximum: 100
drrWeights:
type: object
description: Dynamic Reranker (DRR) weight configuration for optimization
properties:
searchDRRWeight:
type: number
format: float
description: Weight for search DRR signal
minimum: 0
maximum: 5
listingDRRWeight:
type: number
format: float
description: Weight for listing DRR signal
minimum: 0
maximum: 5
clickWeight:
type: number
format: float
description: Weight for click signal
minimum: 0
maximum: 5
buyWeight:
type: number
format: float
description: Weight for buy signal
minimum: 0
maximum: 5
revenueWeight:
type: number
format: float
description: Weight for revenue signal
minimum: 0
maximum: 5
dynamicSynonyms:
type: object
description: Dynamic synonyms configuration
required:
- enabled
- workflowIntervalDays
properties:
enabled:
type: boolean
description: When `true`, dynamic synonyms are enabled
workflowIntervalDays:
type: integer
description: Interval in days between synonym workflow runs
minimum: 7
maximum: 31
example: 14
coldStart:
type: boolean
description: Enables cold start mode for dynamic synonyms. Once enabled, it cannot be disabled. When enabled for the first time, a cold-start-agent resource is created with NOT READY state.
default: false
example: false
autoAcceptThreshold:
type: number
format: float
description: Threshold for automatic synonym acceptance. Must be in [0, 1] range.
minimum: 0
maximum: 1
example: 0.8
dynamicFacets:
type: object
description: Dynamic facets configuration
required:
- enabled
- maxFacets
properties:
enabled:
type: boolean
description: When `true`, dynamic facets are enabled.
maxFacets:
type: integer
description: Maximum number of dynamic facets to return. Must be non-negative.
minimum: 0
example: 10
minCoverageThreshold:
type: number
format: float
description: |
Minimum coverage threshold for a facet to be included. Must be in [0, 1] range.
Coverage is `M/N`, where:
- `M` is the number of occurrences of a facet in the search results.
- `N` is the total number of results.
Higher coverage values mean that a facet is more relevant.
minimum: 0
maximum: 1
example: 0.1
objective:
type: string
description: Objective used to select dynamic facets.
enum:
- coverage
- popularity
example: coverage
geoSearch:
type: object
description: Geo search configuration
properties:
geoSearchableAttribute:
type: string
description: |
Attribute that contains the geolocation of the product.
The attribute's value must be one of the following types:
- JSON object with `lat` and `lon` properties
- JSON array of objects containing `lat` and `lon` properties
- JSON string of format `latitude,longitude`
- JSON array of strings of format `latitude,longitude`
queryRules:
description: Configuration for custom behavior of query rules at the index level.
type: object
properties:
consequencesJoinOperators:
type: object
description: |
Defines how filter consequences from multiple query rules are logically combined.
This setting determines whether those filters are joined using the AND or OR operator.
properties:
standardFilters:
type: string
description: |
Logical operator used to combine standard filters defined in the consequences of multiple query rules.
If not specified, the default operator is OR.
enum:
- AND
- OR
default: OR
elasticFilters:
type: string
description: |
Logical operator used to combine elastic filters defined in the consequences of multiple query rules.
If not specified, the default operator is OR.
enum:
- AND
- OR
default: OR
updatedAt:
type: string
format: date-time
description: Last update time in YYYY-MM-DDThh:mm:ssZ format (ISO 8601, UTC)
createdAt:
type: string
format: date-time
description: Creation time in YYYY-MM-DDThh:mm:ssZ format (ISO 8601, UTC)
indexType:
type: string
description: |
Purpose of the index. Defaults to `Product` when absent. `Document` enables RAG-style chunked article indexing, where each item's body field is split into passages and indexed as separate documents.
Immutable after creation: an update that changes `indexType` (including switching a `Document` index back to `Product` by omitting the field) is rejected. To change the type, create a new index.
enum:
- Product
- Document
chunking:
type: object
description: |
Controls how a `Document` index splits each article body into passages. Only meaningful when `indexType=Document`; ignored for `Product` indices. All fields are optional — sensible defaults are applied when omitted.
properties:
strategy:
type: string
enum:
- section
default: section
description: |
Chunk strategy. `section` = heading-aware sections with nested passages (recommended for RAG).
tokenLimit:
type: integer
minimum: 1
default: 512
description: Target maximum tokens per chunk. An internal safety ratio keeps chunks comfortably under this budget.
overlapTokens:
type: integer
minimum: 0
default: 64
description: Tokens of the previous chunk prepended to the next, preserving cross-chunk context for retrieval.
minChunkTokens:
type: integer
minimum: 1
default: 16
description: Trailing chunks smaller than this fold into the preceding chunk to avoid sub-paragraph noise.
documentFields:
type: object
description: |
Field mapping for `Document` indices. Required when `indexType=Document`.
Best practice: keep `searchableAttributes` minimal — typically `[title]` only. The article body named by `contentField` must NOT be listed in `searchableAttributes`; doing so duplicates the whole body into the parent search tiers on every chunk and is rejected with a 400.
required:
- contentField
properties:
contentField:
type: string
description: |
Item attribute holding the article body to be chunked. Dotted paths are supported for nested fields (e.g. `article.body`). Must reference an existing catalog attribute and must not appear in `searchableAttributes`.
titleField:
type: string
description: Item attribute holding the article title, denormalized onto every chunk for display and ranking. Must exist in the catalog.
contentFormat:
type: string
enum:
- Plaintext
- Markdown
- Html
default: Plaintext
description: How the body is parsed before chunking. `Markdown` and `Html` preserve structure; `Plaintext` splits on paragraph/sentence boundaries.
metadataFields:
type: array
description: Additional catalog attributes denormalized onto every chunk (e.g. author, publishedAt). Each must exist in the catalog.
items:
type: string
"500":
description: Service not available
content:
application/json:
schema:
type: object
properties:
timestamp:
type: string
format: date-time
description: Time when the error occurred
status:
type: integer
description: Status code
error:
type: string
description: Summary of the error
message:
type: string
description: Description of the problem
path:
type: string
description: URL of the requested resource
required:
- timestamp
- status
- message
x-snr-doc-urls:
- /api-reference/ai-search#tag/Search-Configuration/operation/duplicateIndexV2Get
x-codeSamples:
- lang: cURL
label: cURL
source: |-
curl --request GET \
--url https://api.synerise.com/search/v2/indices/%7BindexId%7D/duplicate \
--header 'Authorization: Bearer REPLACE_BEARER_TOKEN'
- lang: Python
label: Python
source: |-
import http.client
conn = http.client.HTTPSConnection("api.synerise.com")
headers = { 'Authorization': "Bearer REPLACE_BEARER_TOKEN" }
conn.request("GET", "/search/v2/indices/%7BindexId%7D/duplicate", headers=headers)
res = conn.getresponse()
data = res.read()
print(data.decode("utf-8"))
- lang: JavaScript
label: JavaScript
source: |-
const data = null;
const xhr = new XMLHttpRequest();
xhr.withCredentials = true;
xhr.addEventListener("readystatechange", function () {
if (this.readyState === this.DONE) {
console.log(this.responseText);
}
});
xhr.open("GET", "https://api.synerise.com/search/v2/indices/%7BindexId%7D/duplicate");
xhr.setRequestHeader("Authorization", "Bearer REPLACE_BEARER_TOKEN");
xhr.send(data);
- lang: Node.js
label: Node.js
source: |-
const http = require("https");
const options = {
"method": "GET",
"hostname": "api.synerise.com",
"port": null,
"path": "/search/v2/indices/%7BindexId%7D/duplicate",
"headers": {
"Authorization": "Bearer REPLACE_BEARER_TOKEN"
}
};
const req = http.request(options, function (res) {
const chunks = [];
res.on("data", function (chunk) {
chunks.push(chunk);
});
res.on("end", function () {
const body = Buffer.concat(chunks);
console.log(body.toString());
});
});
req.end();
- lang: PHP
label: PHP
source: |-
setUrl('https://api.synerise.com/search/v2/indices/%7BindexId%7D/duplicate');
$request->setMethod(HTTP_METH_GET);
$request->setHeaders([
'Authorization' => 'Bearer REPLACE_BEARER_TOKEN'
]);
try {
$response = $request->send();
echo $response->getBody();
} catch (HttpException $ex) {
echo $ex;
}
- lang: Java
label: Java
source: |-
HttpResponse response = Unirest.get("https://api.synerise.com/search/v2/indices/%7BindexId%7D/duplicate")
.header("Authorization", "Bearer REPLACE_BEARER_TOKEN")
.asString();
servers:
- description: Microsoft Azure EU
url: https://api.synerise.com
- description: Microsoft Azure USA
url: https://api.azu.synerise.com
- description: Google Cloud Platform
url: https://api.geb.synerise.com
tags:
- name: Search Configuration
components:
securitySchemes:
JWT:
type: http
scheme: bearer
description: |-
JWT Bearer token. The header looks like this: `Bearer {JWT}`
Remember to include the space between 'Bearer' and the token.
Generate a token via the **Authorization** endpoints.
```