> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.talqora.com/get-started/sdks/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.talqora.com/_mcp/server.

# SDKs

Talqora Vector ships official Python and TypeScript SDKs generated from the same OpenAPI contract as the API Reference.

## Install

**`Python`**

```bash Python
pip install talqora
```

**`TypeScript`**

```bash TypeScript
npm install talqora
```

## Authenticate

Both SDKs read `TALQORA_API_KEY` automatically. You can also pass the API key directly to the client.

**`Python`**

```python Python
import os
from talqora import TalqoraApi

client = TalqoraApi(
    talqora_api_key=os.environ["TALQORA_API_KEY"],
)
```

**`TypeScript`**

```typescript TypeScript
import { TalqoraApiClient } from "talqora";

const client = new TalqoraApiClient({
  talqoraApiKey: process.env.TALQORA_API_KEY,
});
```

## Query vectors

SDK methods are grouped by product resource. You do not need to know the HTTP path or generated OpenAPI operation ID.

**`Python`**

```python Python
dense = client.vectors.query(
    INDEX_ID,
    search_type="dense",
    vector=query_vector,
    top_k=5,
)

sparse = client.vectors.query(
    INDEX_ID,
    search_type="sparse",
    sparse_query="hybrid search",
    top_k=5,
)

hybrid = client.vectors.query(
    INDEX_ID,
    search_type="hybrid",
    vector=query_vector,
    sparse_query="enterprise documents",
    top_k=5,
)
```

**`TypeScript`**

```typescript TypeScript
const hybrid = await client.vectors.query({
  index_id: INDEX_ID,
  search_type: "hybrid",
  vector: queryVector,
  sparse_query: "enterprise documents",
  top_k: 5,
});
```

The same resource-oriented convention applies throughout the SDK: `client.indexes.create(...)`, `client.vectors.upsert(...)`, `client.files.create_upload(...)` in Python, and the corresponding camel-cased methods in TypeScript.

## What is included

* Organization onboarding, members, plans, API keys, and usage.
* Regional indexes with immutable dimensions and distance metrics.
* Dense vector writes, deletes, metadata filters, and queries.
* Sparse BM25 and hybrid retrieval through the same API surface.
* Asynchronous file-processing jobs for extraction, OCR, chunking, and embeddings.
* Connector sync jobs, branches, assistants, and webhooks where enabled.

## Versioning and support

The SDKs are generated from the versioned OpenAPI definition. Pin the package version in production and use the API Reference for endpoint-level request and response details.

* Python package: [pypi.org/project/talqora](https://pypi.org/project/talqora/)
* TypeScript package: [npmjs.com/package/talqora](https://www.npmjs.com/package/talqora)
* API Reference: [docs.talqora.com](https://docs.talqora.com)