--- title: vectors | CodeWeaver Docs description: API reference for codeweaver.providers.types.vectors url: "https://docs.knitli.com/api/providers/types/vectors" type: static generatedAt: "2026-04-17T17:21:09.408Z" --- # vectors [Open in ChatGPT](https://chatgpt.com/?q=Read%20https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F.%20I%20want%20to%20ask%20questions%20about%20it.)[Open in Claude](https://claude.ai/new?q=Read%20https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F.%20I%20want%20to%20ask%20questions%20about%20it.)[View in Markdown](/codeweaver/api/providers/types/vectors.md) [Share on LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F)[Share on X](https://x.com/intent/tweet?url=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F&text=vectors)[Share on Threads](https://threads.net/intent/post?url=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F&text=vectors)[Share on Bluesky](https://bsky.app/intent/compose?text=vectors%20https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F)[Share on Facebook](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F)[Share on Reddit](https://reddit.com/submit?url=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F&title=vectors)[Share on Hacker News](https://news.ycombinator.com/submitlink?u=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F&t=vectors)[Share on Email](mailto:?subject=vectors&body=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F)[Share on WhatsApp](https://wa.me/?text=vectors%20https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F)[Share on Telegram](https://t.me/share/url?url=https%3A%2F%2Fdocs.knitli.com%2Fcodeweaver%2Fapi%2Fproviders%2Ftypes%2Fvectors%2F&text=vectors) # `codeweaver.providers.types.vectors` [Section titled “codeweaver.providers.types.vectors”](#codeweaverproviderstypesvectors) Vector configuration types for Qdrant integration. ## Class: `VectorConfig` [Section titled “Class: VectorConfig”](#class-vectorconfig) Immutable configuration for a single named vector in Qdrant. Aligns directly with Qdrant’s vector configuration while adding minimal metadata for generation and identification. Attributes: name: Physical Qdrant vector name (e.g., “primary”, “backup”, “sparse”) Should reflect the vector’s purpose/role, not the model. model_name: Model used to generate this vector (metadata for generation) params: Direct Qdrant VectorParams or SparseVectorParams role: Optional semantic role (defaults to name if not provided) Example: >>> config = VectorConfig( … name=“primary”, … model_name=ModelName(“voyage-code-3”), … params=VectorParams(size=1024, distance=“Cosine”), … role=VectorRole.PRIMARY, … ) ### Method: `from_provider_settings` [Section titled “Method: from_provider_settings”](#method-from_provider_settings) ** ``` from_provider_settings() ``` Create VectorConfig from provider settings. Args: config: Provider settings (embedding or sparse) name: Physical Qdrant vector name (REQUIRED, purpose-based) e.g., “primary”, “backup”, “sparse”, “semantic_code” role: Optional semantic role (defaults to name) Returns: Configured VectorConfig instance Example: >>> from codeweaver.providers.config.categories import EmbeddingProviderSettings >>> settings = EmbeddingProviderSettings(…) >>> config = await VectorConfig.from_provider_settings( … settings, name=“primary”, role=VectorRole.PRIMARY … ) ### Method: `to_qdrant_config` [Section titled “Method: to_qdrant_config”](#method-to_qdrant_config) ** ``` to_qdrant_config() ``` Return (name, params) tuple for Qdrant collection config. Returns: Tuple of (physical_vector_name, vector_params) ready for Qdrant’s vectors_config or sparse_vectors_config dict. Example: >>> config = VectorConfig(name=“primary”, …) >>> name, params = config.to_qdrant_config() >>> collection_config = CollectionParams(vectors_config={name: params}) ## Class: `VectorRole` [Section titled “Class: VectorRole”](#class-vectorrole) Semantic roles for vectors (extensible for future strategies). The enum values should match common physical vector names by default. Use .variable property to get snake_case string representation. Example: >>> VectorRole.PRIMARY.variable ‘primary’ >>> VectorRole.BACKUP.variable ‘backup’ Future extensions might include: - SEMANTIC_CODE: “semantic_code” - SEMANTIC_DOCS: “semantic_docs” - KEYWORD: “keyword” ## Class: `VectorSet` [Section titled “Class: VectorSet”](#class-vectorset) Immutable collection of vectors for a search/indexing strategy. Manages multiple named vectors with flexible organization. Vectors can be keyed by strategy name, intent, or any identifier. Provides query methods and conversion to Qdrant collection configuration. Attributes: vectors: Dict of VectorConfig keyed by logical identifier (e.g., “primary”, “backup”, “sparse”, “semantic_v1”) Example: >>> vector_set = VectorSet(vectors={ … “primary”: VectorConfig(name=“primary”, …), … “backup”: VectorConfig(name=“backup”, …), … “sparse”: VectorConfig(name=“sparse”, …), … }) >>> >>> # Convert to Qdrant config >>> vectors_config = vector_set.to_qdrant_vectors_config() >>> sparse_config = vector_set.to_qdrant_sparse_vectors_config() ### Method: `backup` [Section titled “Method: backup”](#method-backup) ** ``` backup() ``` Get backup dense vector (common case). Returns first vector with BACKUP role, or None. ### Method: `by_key` [Section titled “Method: by_key”](#method-by_key) ** ``` by_key() ``` Get vector by its logical dict key. Args: key: Logical identifier used in vectors dict Returns: Matching VectorConfig or None ### Method: `by_name` [Section titled “Method: by_name”](#method-by_name) ** ``` by_name() ``` Get vector by its physical Qdrant name. Args: name: Physical vector name (e.g., “primary”, “backup”) Returns: Matching VectorConfig or None ### Method: `by_role` [Section titled “Method: by_role”](#method-by_role) ** ``` by_role() ``` Get all vectors with a specific role. Supports multiple vectors with the same role for future multi-strategy support. Args: role: VectorRole enum or string role name Returns: List of matching VectorConfigs (may be empty) Example: >>> primaries = vector_set.by_role(VectorRole.PRIMARY) >>> backups = vector_set.by_role(“backup”) ### Method: `default` [Section titled “Method: default”](#method-default) ** ``` default() ``` Create default VectorSet with recommended configuration. Returns: VectorSet with recommended profile configuration ### Method: `dense_vectors` [Section titled “Method: dense_vectors”](#method-dense_vectors) ** ``` dense_vectors() ``` Get all dense vectors. Returns: Dict of dense VectorConfigs (preserves keys) ### Method: `from_profile` [Section titled “Method: from_profile”](#method-from_profile) ** ``` from_profile() ``` Create VectorSet from provider profile. Creates standard layout with primary, backup (optional), and sparse (optional) vectors based on profile configuration. Uses role-based physical vector names. Args: profile: Provider profile with embedding configurations Returns: Configured VectorSet with standard layout Example: >>> from codeweaver.providers.config.profiles import ProviderProfile >>> vector_set = await VectorSet.from_profile(ProviderProfile.RECOMMENDED) ### Method: `primary` [Section titled “Method: primary”](#method-primary) ** ``` primary() ``` Get primary dense vector (common case). Returns first vector with PRIMARY role, or None. ### Method: `sparse` [Section titled “Method: sparse”](#method-sparse) ** ``` sparse() ``` Get sparse vector (common case). Returns first vector with SPARSE role, or None. ### Method: `sparse_vectors` [Section titled “Method: sparse_vectors”](#method-sparse_vectors) ** ``` sparse_vectors() ``` Get all sparse vectors. Returns: Dict of SparseVectorConfigs (preserves keys) ### Method: `to_qdrant_sparse_vectors_config` [Section titled “Method: to_qdrant_sparse_vectors_config”](#method-to_qdrant_sparse_vectors_config) ** ``` to_qdrant_sparse_vectors_config() ``` Generate Qdrant sparse_vectors_config dict for sparse vectors. Returns: Dict mapping physical vector names to SparseVectorParams, ready for CollectionConfig.sparse_vectors_config Example: >>> config = CollectionParams( … sparse_vectors_config=vector_set.to_qdrant_sparse_vectors_config() … ) ### Method: `to_qdrant_vectors_config` [Section titled “Method: to_qdrant_vectors_config”](#method-to_qdrant_vectors_config) ** ``` to_qdrant_vectors_config() ``` Generate Qdrant vectors_config dict for dense vectors. Returns: Dict mapping physical vector names to VectorParams, ready for CollectionConfig.vectors_config Example: >>> config = CollectionParams(vectors_config=vector_set.to_qdrant_vectors_config())