Search Index, REST: SearchIndex.Create

Written by

Updated at December 17, 2024

HTTP request
Body parameters
ExpirationConfig
TextSearchIndex
ChunkingStrategy
StaticChunkingStrategy
VectorSearchIndex
HybridSearchIndex
CombinationStrategy
MeanCombinationStrategy
ReciprocalRankFusionCombinationStrategy
Response
Status
SearchIndex
ExpirationConfig
TextSearchIndex
ChunkingStrategy
StaticChunkingStrategy
VectorSearchIndex
HybridSearchIndex
CombinationStrategy
MeanCombinationStrategy
ReciprocalRankFusionCombinationStrategy

Create a new search index in asynchronous mode.

HTTP request

POST https://rest-assistant.api.cloud.yandex.net/assistants/v1/searchIndex

Body parameters

{
  "folderId": "string",
  "fileIds": [
    "string"
  ],
  "name": "string",
  "description": "string",
  "expirationConfig": {
    "expirationPolicy": "string",
    "ttlDays": "string"
  },
  "labels": "object",
  // Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`
  "textSearchIndex": {
    "chunkingStrategy": {
      // Includes only one of the fields `staticStrategy`
      "staticStrategy": {
        "maxChunkSizeTokens": "string",
        "chunkOverlapTokens": "string"
      }
      // end of the list of possible fields
    }
  },
  "vectorSearchIndex": {
    "docEmbedderUri": "string",
    "queryEmbedderUri": "string",
    "chunkingStrategy": {
      // Includes only one of the fields `staticStrategy`
      "staticStrategy": {
        "maxChunkSizeTokens": "string",
        "chunkOverlapTokens": "string"
      }
      // end of the list of possible fields
    }
  },
  "hybridSearchIndex": {
    "textSearchIndex": {
      "chunkingStrategy": {
        // Includes only one of the fields `staticStrategy`
        "staticStrategy": {
          "maxChunkSizeTokens": "string",
          "chunkOverlapTokens": "string"
        }
        // end of the list of possible fields
      }
    },
    "vectorSearchIndex": {
      "docEmbedderUri": "string",
      "queryEmbedderUri": "string",
      "chunkingStrategy": {
        // Includes only one of the fields `staticStrategy`
        "staticStrategy": {
          "maxChunkSizeTokens": "string",
          "chunkOverlapTokens": "string"
        }
        // end of the list of possible fields
      }
    },
    "chunkingStrategy": {
      // Includes only one of the fields `staticStrategy`
      "staticStrategy": {
        "maxChunkSizeTokens": "string",
        "chunkOverlapTokens": "string"
      }
      // end of the list of possible fields
    },
    "normalizationStrategy": "string",
    "combinationStrategy": {
      // Includes only one of the fields `meanCombination`, `rrfCombination`
      "meanCombination": {
        "meanEvaluationTechnique": "string",
        "weights": [
          "string"
        ]
      },
      "rrfCombination": {
        "k": "string"
      }
      // end of the list of possible fields
    }
  }
  // end of the list of possible fields
}

Request to create a new search index.

Field	Description
folderId	string Required field.
fileIds[]	string List of file IDs to be indexed.
name	string Name of the search index.
description	string Description of the search index.
expirationConfig	ExpirationConfig Expiration configuration for the search index.
labels	object (map<string, string>) Set of key-value pairs to label the search index.
textSearchIndex	TextSearchIndex Configuration for a traditional keyword-based text search index. Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`.
vectorSearchIndex	VectorSearchIndex Configuration for a vector-based search index using embeddings. Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`.
hybridSearchIndex	HybridSearchIndex Configuration for a hybrid (vector-based + keyword-based) search index. Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`.

ExpirationConfig

Field

Description

expirationPolicy

enum (ExpirationPolicy)

EXPIRATION_POLICY_UNSPECIFIED
STATIC
SINCE_LAST_ACTIVE

ttlDays

string (int64)

TextSearchIndex

Defines the configuration for a traditional keyword-based text search index.

Field

Description

chunkingStrategy

ChunkingStrategy

Chunking strategy used to split text into smaller chunks before indexing.
In the case of text search, tokens are individual text characters.

ChunkingStrategy

Defines a general strategy for chunking text into smaller segments.
Currently, only StaticChunkingStrategy is supported.

Field

Description

staticStrategy

StaticChunkingStrategy

Includes only one of the fields staticStrategy.

StaticChunkingStrategy

Defines a chunking strategy where chunks are created with a fixed maximum chunk size and an overlap between consecutive chunks.

Field

Description

maxChunkSizeTokens

string (int64)

The maximum number of tokens allowed in a single chunk.
Constraints: must be within the range [100, 2048].
Default value: 800

chunkOverlapTokens

string (int64)

The number of tokens that should overlap between consecutive chunks.
This allows for some context from the previous chunk to be included in the next chunk.
Constraints: must be less than or equal to half of max_chunk_size_tokens.
Default value: 400

VectorSearchIndex

Defines the configuration for a vector-based search index. This type uses embeddings to represent documents and queries.

Field	Description
docEmbedderUri	string The ID of the model to be used for obtaining document text embeddings.
queryEmbedderUri	string The ID of the model to be used for obtaining query text embeddings.
chunkingStrategy	ChunkingStrategy Chunking strategy used to split text into smaller chunks before indexing. In the case of vector search, tokens are produced by the tokenizer from the embedding model.

HybridSearchIndex

Defines the configuration for a hybrid (vector-based + keyword-based) search index. This type uses both embeddings and keyword-based search to represent documents and queries.

Field	Description
textSearchIndex	TextSearchIndex Configuration for a traditional keyword-based text search index.
vectorSearchIndex	VectorSearchIndex Configuration for a vector-based search index.
chunkingStrategy	ChunkingStrategy Common chunking strategy that applies to both text and vector search indexes. If provided, it overrides the individual chunking strategies in both `text_search_index` and `vector_search_index`. In this case, both text and vector search will use token-based chunking, where tokens are produced by the tokenizer of the embedding model.
normalizationStrategy	enum (NormalizationStrategy) Normalization strategy for relevance scores from different indices. Default is MIN_MAX_STRATEGY `NORMALIZATION_STRATEGY_UNSPECIFIED` `MIN_MAX`: https://en.wikipedia.org/wiki/Feature_scaling#Rescaling_(min-max_normalization) `L2`: https://en.wikipedia.org/wiki/Cosine_similarity#L2-normalized_Euclidean_distance
combinationStrategy	CombinationStrategy Combination strategy for merging rankings from different indices. Default is arithmetic mean

CombinationStrategy

Combination strategy for merging rankings from different indices

Field

Description

meanCombination

MeanCombinationStrategy

Includes only one of the fields meanCombination, rrfCombination.

rrfCombination

ReciprocalRankFusionCombinationStrategy

Includes only one of the fields meanCombination, rrfCombination.

MeanCombinationStrategy

Field

Description

meanEvaluationTechnique

enum (MeanEvaluationTechnique)

Technique for averaging relevance scores from different indices. Default is ARITHMETIC

MEAN_EVALUATION_TECHNIQUE_UNSPECIFIED
ARITHMETIC: https://en.wikipedia.org/wiki/Arithmetic_mean
GEOMETRIC: https://en.wikipedia.org/wiki/Geometric_mean
HARMONIC: https://en.wikipedia.org/wiki/Harmonic_mean

weights[]

string

Weights used for evaluating the weighted mean of relevance scores. The sum of the values must equal 1.0
If not provided, all scores are given equal weight

ReciprocalRankFusionCombinationStrategy

https://plg.uwaterloo.ca/~gvcormac/cormacksigir09-rrf.pdf

Field

Description

string (int64)

The parameter k for RRFscore. Default is 60

Response

HTTP Code: 200 - OK

{
  "id": "string",
  "description": "string",
  "createdAt": "string",
  "createdBy": "string",
  "modifiedAt": "string",
  "done": "boolean",
  "metadata": "object",
  // Includes only one of the fields `error`, `response`
  "error": {
    "code": "integer",
    "message": "string",
    "details": [
      "object"
    ]
  },
  "response": {
    "id": "string",
    "folderId": "string",
    "name": "string",
    "description": "string",
    "createdBy": "string",
    "createdAt": "string",
    "updatedBy": "string",
    "updatedAt": "string",
    "expirationConfig": {
      "expirationPolicy": "string",
      "ttlDays": "string"
    },
    "expiresAt": "string",
    "labels": "object",
    // Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`
    "textSearchIndex": {
      "chunkingStrategy": {
        // Includes only one of the fields `staticStrategy`
        "staticStrategy": {
          "maxChunkSizeTokens": "string",
          "chunkOverlapTokens": "string"
        }
        // end of the list of possible fields
      }
    },
    "vectorSearchIndex": {
      "docEmbedderUri": "string",
      "queryEmbedderUri": "string",
      "chunkingStrategy": {
        // Includes only one of the fields `staticStrategy`
        "staticStrategy": {
          "maxChunkSizeTokens": "string",
          "chunkOverlapTokens": "string"
        }
        // end of the list of possible fields
      }
    },
    "hybridSearchIndex": {
      "textSearchIndex": {
        "chunkingStrategy": {
          // Includes only one of the fields `staticStrategy`
          "staticStrategy": {
            "maxChunkSizeTokens": "string",
            "chunkOverlapTokens": "string"
          }
          // end of the list of possible fields
        }
      },
      "vectorSearchIndex": {
        "docEmbedderUri": "string",
        "queryEmbedderUri": "string",
        "chunkingStrategy": {
          // Includes only one of the fields `staticStrategy`
          "staticStrategy": {
            "maxChunkSizeTokens": "string",
            "chunkOverlapTokens": "string"
          }
          // end of the list of possible fields
        }
      },
      "chunkingStrategy": {
        // Includes only one of the fields `staticStrategy`
        "staticStrategy": {
          "maxChunkSizeTokens": "string",
          "chunkOverlapTokens": "string"
        }
        // end of the list of possible fields
      },
      "normalizationStrategy": "string",
      "combinationStrategy": {
        // Includes only one of the fields `meanCombination`, `rrfCombination`
        "meanCombination": {
          "meanEvaluationTechnique": "string",
          "weights": [
            "string"
          ]
        },
        "rrfCombination": {
          "k": "string"
        }
        // end of the list of possible fields
      }
    }
    // end of the list of possible fields
  }
  // end of the list of possible fields
}

An Operation resource. For more information, see Operation.

Field	Description
id	string ID of the operation.
description	string Description of the operation. 0-256 characters long.
createdAt	string (date-time) Creation timestamp. String in RFC3339 text format. The range of possible values is from `0001-01-01T00:00:00Z` to `9999-12-31T23:59:59.999999999Z`, i.e. from 0 to 9 digits for fractions of a second. To work with values in this field, use the APIs described in the Protocol Buffers reference. In some languages, built-in datetime utilities do not support nanosecond precision (9 digits).
createdBy	string ID of the user or service account who initiated the operation.
modifiedAt	string (date-time) The time when the Operation resource was last modified. String in RFC3339 text format. The range of possible values is from `0001-01-01T00:00:00Z` to `9999-12-31T23:59:59.999999999Z`, i.e. from 0 to 9 digits for fractions of a second. To work with values in this field, use the APIs described in the Protocol Buffers reference. In some languages, built-in datetime utilities do not support nanosecond precision (9 digits).
done	boolean If the value is `false`, it means the operation is still in progress. If `true`, the operation is completed, and either `error` or `response` is available.
metadata	object Service-specific metadata associated with the operation. It typically contains the ID of the target resource that the operation is performed on. Any method that returns a long-running operation should document the metadata type, if any.
error	Status The error result of the operation in case of failure or cancellation. Includes only one of the fields `error`, `response`. The operation result. If `done == false` and there was no failure detected, neither `error` nor `response` is set. If `done == false` and there was a failure detected, `error` is set. If `done == true`, exactly one of `error` or `response` is set.
response	SearchIndex The normal response of the operation in case of success. If the original method returns no data on success, such as Delete, the response is google.protobuf.Empty. If the original method is the standard Create/Update, the response should be the target resource of the operation. Any method that returns a long-running operation should document the response type, if any. Includes only one of the fields `error`, `response`. The operation result. If `done == false` and there was no failure detected, neither `error` nor `response` is set. If `done == false` and there was a failure detected, `error` is set. If `done == true`, exactly one of `error` or `response` is set.

Status

The error result of the operation in case of failure or cancellation.

Field	Description
code	integer (int32) Error code. An enum value of google.rpc.Code.
message	string An error message.
details[]	object A list of messages that carry the error details.

SearchIndex

Represents a search index used to store and query data, either using traditional keyword-based text search or vector-based search mechanisms.

Field	Description
id	string Unique identifier of the search index.
folderId	string ID of the folder that the search index belongs to.
name	string Name of the search index.
description	string Description of the search index.
createdBy	string Identifier of the subject who created this search index.
createdAt	string (date-time) Timestamp representing when the search index was created. String in RFC3339 text format. The range of possible values is from `0001-01-01T00:00:00Z` to `9999-12-31T23:59:59.999999999Z`, i.e. from 0 to 9 digits for fractions of a second. To work with values in this field, use the APIs described in the Protocol Buffers reference. In some languages, built-in datetime utilities do not support nanosecond precision (9 digits).
updatedBy	string Identifier of the subject who last updated this search index.
updatedAt	string (date-time) Timestamp representing the last time this search index was updated. String in RFC3339 text format. The range of possible values is from `0001-01-01T00:00:00Z` to `9999-12-31T23:59:59.999999999Z`, i.e. from 0 to 9 digits for fractions of a second. To work with values in this field, use the APIs described in the Protocol Buffers reference. In some languages, built-in datetime utilities do not support nanosecond precision (9 digits).
expirationConfig	ExpirationConfig Configuration for the expiration of the search index, defining when and how the search index will expire.
expiresAt	string (date-time) Timestamp representing when the search index will expire. String in RFC3339 text format. The range of possible values is from `0001-01-01T00:00:00Z` to `9999-12-31T23:59:59.999999999Z`, i.e. from 0 to 9 digits for fractions of a second. To work with values in this field, use the APIs described in the Protocol Buffers reference. In some languages, built-in datetime utilities do not support nanosecond precision (9 digits).
labels	object (map<string, string>) Set of key-value pairs that can be used to organize and categorize the search index.
textSearchIndex	TextSearchIndex Keyword-based text search index configuration. This type of index is used for traditional text search, where documents are indexed based on their keywords. Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`. Type of the search index. It can be either a traditional keyword-based text search or a vector-based search.
vectorSearchIndex	VectorSearchIndex Vector-based search index configuration. This type is used for vector search, where documents are indexed using vector embeddings. Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`. Type of the search index. It can be either a traditional keyword-based text search or a vector-based search.
hybridSearchIndex	HybridSearchIndex Hybrid (vector-based + keyword-based) search index configuration This type is used for hybrid search, where documents are indexed using both keyword-based and vector-based search mechanisms. Includes only one of the fields `textSearchIndex`, `vectorSearchIndex`, `hybridSearchIndex`. Type of the search index. It can be either a traditional keyword-based text search or a vector-based search.

ExpirationConfig

Field

Description

expirationPolicy

enum (ExpirationPolicy)

EXPIRATION_POLICY_UNSPECIFIED
STATIC
SINCE_LAST_ACTIVE

ttlDays

string (int64)

TextSearchIndex

Defines the configuration for a traditional keyword-based text search index.

Field

Description

chunkingStrategy

ChunkingStrategy

Chunking strategy used to split text into smaller chunks before indexing.
In the case of text search, tokens are individual text characters.

ChunkingStrategy

Defines a general strategy for chunking text into smaller segments.
Currently, only StaticChunkingStrategy is supported.

Field

Description

staticStrategy

StaticChunkingStrategy

Includes only one of the fields staticStrategy.

StaticChunkingStrategy

Defines a chunking strategy where chunks are created with a fixed maximum chunk size and an overlap between consecutive chunks.

Field

Description

maxChunkSizeTokens

string (int64)

The maximum number of tokens allowed in a single chunk.
Constraints: must be within the range [100, 2048].
Default value: 800

chunkOverlapTokens

string (int64)

VectorSearchIndex

Defines the configuration for a vector-based search index. This type uses embeddings to represent documents and queries.

Field	Description
docEmbedderUri	string The ID of the model to be used for obtaining document text embeddings.
queryEmbedderUri	string The ID of the model to be used for obtaining query text embeddings.
chunkingStrategy	ChunkingStrategy Chunking strategy used to split text into smaller chunks before indexing. In the case of vector search, tokens are produced by the tokenizer from the embedding model.

HybridSearchIndex

Defines the configuration for a hybrid (vector-based + keyword-based) search index. This type uses both embeddings and keyword-based search to represent documents and queries.

Field	Description
textSearchIndex	TextSearchIndex Configuration for a traditional keyword-based text search index.
vectorSearchIndex	VectorSearchIndex Configuration for a vector-based search index.
chunkingStrategy	ChunkingStrategy Common chunking strategy that applies to both text and vector search indexes. If provided, it overrides the individual chunking strategies in both `text_search_index` and `vector_search_index`. In this case, both text and vector search will use token-based chunking, where tokens are produced by the tokenizer of the embedding model.
normalizationStrategy	enum (NormalizationStrategy) Normalization strategy for relevance scores from different indices. Default is MIN_MAX_STRATEGY `NORMALIZATION_STRATEGY_UNSPECIFIED` `MIN_MAX`: https://en.wikipedia.org/wiki/Feature_scaling#Rescaling_(min-max_normalization) `L2`: https://en.wikipedia.org/wiki/Cosine_similarity#L2-normalized_Euclidean_distance
combinationStrategy	CombinationStrategy Combination strategy for merging rankings from different indices. Default is arithmetic mean

CombinationStrategy

Combination strategy for merging rankings from different indices

Field

Description

meanCombination

MeanCombinationStrategy

Includes only one of the fields meanCombination, rrfCombination.

rrfCombination

ReciprocalRankFusionCombinationStrategy

Includes only one of the fields meanCombination, rrfCombination.

MeanCombinationStrategy

Field

Description

meanEvaluationTechnique

enum (MeanEvaluationTechnique)

Technique for averaging relevance scores from different indices. Default is ARITHMETIC

MEAN_EVALUATION_TECHNIQUE_UNSPECIFIED
ARITHMETIC: https://en.wikipedia.org/wiki/Arithmetic_mean
GEOMETRIC: https://en.wikipedia.org/wiki/Geometric_mean
HARMONIC: https://en.wikipedia.org/wiki/Harmonic_mean

weights[]

string

Weights used for evaluating the weighted mean of relevance scores. The sum of the values must equal 1.0
If not provided, all scores are given equal weight

ReciprocalRankFusionCombinationStrategy

https://plg.uwaterloo.ca/~gvcormac/cormacksigir09-rrf.pdf

Field

Description

string (int64)

The parameter k for RRFscore. Default is 60

Search Index, REST: SearchIndex.Create

HTTP requestHTTP request

Body parametersBody parameters

ExpirationConfigExpirationConfig

TextSearchIndexTextSearchIndex

ChunkingStrategyChunkingStrategy

StaticChunkingStrategyStaticChunkingStrategy

VectorSearchIndexVectorSearchIndex

HybridSearchIndexHybridSearchIndex

CombinationStrategyCombinationStrategy

MeanCombinationStrategyMeanCombinationStrategy

ReciprocalRankFusionCombinationStrategyReciprocalRankFusionCombinationStrategy

ResponseResponse

StatusStatus

SearchIndexSearchIndex

ExpirationConfigExpirationConfig

TextSearchIndexTextSearchIndex

ChunkingStrategyChunkingStrategy

StaticChunkingStrategyStaticChunkingStrategy

VectorSearchIndexVectorSearchIndex

HybridSearchIndexHybridSearchIndex

CombinationStrategyCombinationStrategy

MeanCombinationStrategyMeanCombinationStrategy

ReciprocalRankFusionCombinationStrategyReciprocalRankFusionCombinationStrategy

Was the article helpful?

HTTP request

Body parameters

ExpirationConfig

TextSearchIndex

ChunkingStrategy

StaticChunkingStrategy

VectorSearchIndex

HybridSearchIndex

CombinationStrategy

MeanCombinationStrategy

ReciprocalRankFusionCombinationStrategy

Response

Status

SearchIndex

ExpirationConfig

TextSearchIndex

ChunkingStrategy

StaticChunkingStrategy

VectorSearchIndex

HybridSearchIndex

CombinationStrategy

MeanCombinationStrategy

ReciprocalRankFusionCombinationStrategy