Tool Integrations
⚙Tool
Qdrant
Store and search vectors in Qdrant
Qdrant
Store, search, and manage vector embeddings in Qdrant — a high-performance open-source vector database. Supports filtering, payload management, and batch operations.
Overview
| Property | Value |
|---|---|
| Type | qdrant |
| Category | Tool — Vector Database |
| Auth | API Key |
Operations
| Operation | Description |
|---|---|
| Upsert | Insert or update vectors with payloads |
| Search | Find similar vectors |
| Delete | Remove vectors |
| List Collections | Get available collections |
Configuration
| Setting | Type | Description |
|---|---|---|
| API Key | Password | Qdrant API key |
| URL | Short input | Qdrant cluster URL |
| Collection | Short input | Collection name |
| Top K | Slider | Number of results |
| Vector | Code editor | Query vector |
| Filter | Code editor | Qdrant filter conditions (JSON) |
Outputs
| Field | Type | Description |
|---|---|---|
results | json | Similar vectors with scores and payloads |
content | string | Formatted results |
Example: Filtered Semantic Search
Workflow:
[Starter: Query + Category] → [OpenAI: Embed] → [Qdrant: Search] → [Agent] → [Response]Qdrant Filter:
{
"must": [
{ "key": "category", "match": { "value": "{{starter.input.category}}" } }
]
}Search similar vectors but only within a specific category — combining semantic similarity with metadata filtering.
Tips
- Payload filters are powerful — combine semantic search with exact match filters
- Self-hosted option — Qdrant can run locally with Docker
- Higher dimension support than some alternatives — works with any embedding model