Skip to main content

LambdaDB Python SDK

Developer-friendly & type-safe Python SDK specifically catered to leverage LambdaDB API.

Summary

LambdaDB API: LambdaDB Open API Spec

Table of Contents

SDK Installation

The SDK currently supports Python >=3.10,<3.14. Python 3.9 users can remain on the stable 0.9.0 release until they upgrade their interpreter.

The SDK can be installed with uv, pip, or poetry package managers.

Before upgrading to 0.13.0, review the vector configuration migration requirements. Dense vectors now follow the public OpenAPI: 64-4096 dimensions and cosine/euclidean similarity, including collection response parsing. Collections outside that contract must be migrated before upgrading.

uv

uv is a fast Python package installer and resolver, designed as a drop-in replacement for pip and pip-tools. It's recommended for its speed and modern Python tooling capabilities.

uv add lambdadb

PIP

PIP is the default package installer for Python, enabling easy installation and management of packages from PyPI via the command line.

pip install lambdadb

Poetry

Poetry is a modern tool that simplifies dependency management and package publishing by using a single pyproject.toml file to handle project metadata and dependencies.

poetry add lambdadb

Shell and script usage with uv

You can use this SDK in a Python shell with uv and the uvx command that comes with it like so:

uvx --from lambdadb python

It's also possible to write a standalone Python script without needing to set up a whole project like so:

#!/usr/bin/env -S uv run --script
# /// script
# requires-python = ">=3.10,<3.14"
# dependencies = [
#     "lambdadb",
# ]
# ///

from lambdadb import LambdaDB

client = LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
)

# Rest of script here...

Once that is saved to a file, you can run it with uv run script.py where script.py can be replaced with the actual file name.

Qdrant Compatibility

LambdaDB includes explicit compatibility clients for migrating common vector database workflows with minimal application changes. The first supported compatibility layer is Qdrant:

from lambdadb.compat.qdrant import QdrantCompatClient, models

client = QdrantCompatClient(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
)

See Qdrant compatibility for supported APIs, payload index behavior, data mapping, limitations, and live test setup.

IDE Support

PyCharm

Generally, the SDK will work well with most IDEs out of the box. However, when using PyCharm, you can enjoy much better integration with Pydantic by installing an additional plugin.

SDK Example Usage

Use base_url and project_name (defaults: https://api.lambdadb.ai, playground) and the collection-scoped API for the best experience:

from lambdadb import LambdaDB

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",  # optional, this is the default
    project_name="playground",            # optional, this is the default
) as client:
    # Collection-scoped: no need to pass collection_name to every call
    coll = client.collection("my_collection")
    list_res = coll.docs.list()
    items = list_res.results          # or list_res.documents for doc bodies only
    res = coll.query(query={"queryString": {"query": "some text"}})
    docs_only = res.documents         # document bodies; use res.results for score/metadata
    coll.docs.upsert(docs=[{"id": "1", "text": "hello"}])

Data Versioning

Use typed helpers to scope reads, writes, and collection-local snapshots:

from lambdadb import AliasTarget, BranchSource, Ref, RefSource

branch = coll.branches.create("experiment", source=BranchSource.branch("main"))
direct_parent = branch.branch.parent_branch  # fixed branch_id/name, or None
coll.tags.create("validated-2026-09", source=RefSource.branch("experiment"))
coll.aliases.create(
    "production-read", target=AliasTarget.tag("validated-2026-09")
)

results = coll.query(
    query={"queryString": {"query": "text:hello"}},
    ref=Ref.alias("production-read"),
)
coll.docs.upsert(docs=[{"id": "2", "text": "draft"}], branch="experiment")

See the Data Versioning SDK guide for sync and async lifecycle, pagination, and signed bulk-upload examples. If upgrading from stable 0.8.2, review the 0.9.0 migration notes for changed Collection models, timestamp units, HTTP statuses, and Branch source rules before upgrading.

Create a collection with native embeddings

Provide embedding to enable native embeddings on a compatible server. The legacy managedEmbedding=True flag remains supported. Put native dimensions/similarity inside embedding; caller-provided vectors keep top-level dimensions. See the native embedding and Bayesian guide for the pinned contract, update examples, and server requirements. Cohere embeddings, ordered text/image sources, knn.queryInputs, and direct vectors on native fields are described in native embedding sources.

from lambdadb import LambdaDB, models

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
) as client:
    client.collections.create(
        collection_name="articles",
        index_configs={
            "body": {
                "type": models.TypeText.TEXT,
                "analyzers": [models.Analyzer.ENGLISH],
            },
            "bodyEmbedding": {
                "type": models.TypeVector.VECTOR,
                "embedding": {
                    "provider": models.Provider.OPENAI,
                    "model": "text-embedding-3-small",
                    "sourceField": "body",
                },
            },
        },
    )

Use exactly two unboosted signals and an explicit candidate budget:

response = client.collection("articles").query(
    query={"bayesian": [
        {"queryString": {"query": "body:restore"}},
        {"knn": {"field": "bodyEmbedding", "queryText": "Restore a saved version", "k": 30}},
    ]},
    size=10,
    candidate_size=30,
)

With native reranking, omit top-level candidate_size and use the rerank budget. See Bayesian search for limits and examples.

Choose text analyzers

Text fields in this source checkout accept the 49 names in the Analyzer reference. For example, select Chinese word segmentation when creating a collection:

client.collections.create(
    collection_name="chinese-articles",
    index_configs={
        "content": models.IndexConfigsText(
            type=models.TypeText.TEXT,
            analyzers=[models.Analyzer.CHINESE],
        ),
    },
)

CHINESE uses Simplified Chinese word segmentation; CJK uses overlapping two-character tokens. Each selected analyzer processes the field independently; language detection is not automatic. Omit analyzers to use the server default ["standard"]. An empty list is sent as-is and does not select that default; duplicates are also passed through, though selecting each name once is recommended. See Choose text analyzers for selection tradeoffs for the original 16 names shipped in 0.10.0. The 0.11.0 release adds 33 presets, pinned to backend PR #437, merge 55d888299fee44466326a9db8016af9811ade13b; using them requires a compatible server. This source revision does not establish deployment in any environment.

Use lowercase string values or uppercase enum members, for example models.Analyzer.WHITESPACE or models.Analyzer.NEPALI. The SDK's existing case-sensitive validation is unchanged; it does not rewrite configured names. All 49 names are fixed presets for text fields. models.Analyzer.KEYWORD keeps the field type text and does not enable keyword-field sorting or facets. Custom pipelines and analyzer options are unsupported. NEPALI, TAMIL, and TELUGU are Lucene extensions, not common Elasticsearch/OpenSearch analyzers.

Native reranking

This source checkout supports optional native reranking on each query. The server runs reranking and handles provider credentials:

response = client.collection("articles").query(
    query={"queryString": {"query": "body:restore"}},
    size=10,
    rerank=models.RerankConfig(
        provider="typesafe",
        model="jev-1.13.0",
        query_text="How do I restore a previous collection version?",
        fields=["title", "body"],
    ),
)
for item in response.results:
    print(item.score, item.retrieval_score, item.doc)

Cohere rerank-v5.0-pro and rerank-v5.0-fast are also supported without criteria. No provider API key is required. Omitted/null reranking preserves existing search behavior. Check response.rerank.status: applied scores are evaluation values in [0, 1], while fallback keeps search scores. The 0.11.0 release requires a compatible server. See native reranking for custom criteria, candidate limits, response metadata and failure handling.

Count keyword facets

The 0.10.0 release supports keyword facets on a compatible server with newly built keyword indexes:

result = client.collection("items").query(size=0, facets={"tags": {"size": 5}})
for bucket in result.facets["tags"].buckets:
    print(bucket.value, bucket.count)

Counts cover all matching documents; total remains the number of returned documents. query_async() accepts the same arguments. Existing data must be rebuilt into a new Collection before using this index format. See keyword facets for query support, limits, and migration.

List all collections (sync / async)

# Synchronous
from lambdadb import LambdaDB

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
) as client:
    res = client.collections.list()
    print(res)
# Asynchronous
import asyncio
from lambdadb import LambdaDB

async def main():
    async with LambdaDB(
        project_api_key="<YOUR_PROJECT_API_KEY>",
        base_url="https://api.lambdadb.ai",
        project_name="playground",
    ) as client:
        res = await client.collections.list_async()
        print(res)

asyncio.run(main())

Authentication

Per-Client Security Schemes

This SDK supports the following security scheme globally:

Name Type Scheme Environment Variable
project_api_key apiKey API key LAMBDADB_PROJECT_API_KEY

To authenticate with the API the project_api_key parameter must be set when initializing the SDK client instance. Use base_url and project_name for the API endpoint (defaults: https://api.lambdadb.ai, playground). For example:

from lambdadb import LambdaDB

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
) as client:
    res = client.collections.list()
    print(res)

Vector configuration

Dense vector fields require dimensions 64-4096 and cosine or euclidean similarity, following the public OpenAPI contract. These constraints also apply to native embedding configuration and collection response parsing. Review the linked upgrade requirements for legacy schemas.

Available Resources and Operations

Recommended: Use the collection-scoped API: client.collection("name").docs.list(), .docs.fetch(), .docs.upsert(), etc., and client.collection("name").query() for search. This matches the REST API structure and avoids repeating the collection name.

  • Response access: List, query, and fetch responses expose .results (full result items, with score/metadata when applicable) and .documents (document bodies only). When the API returns is_docs_inline: false with a presigned docs_url, the SDK automatically fetches from that URL so response.results and response.documents are always populated when using coll.query() and coll.docs.fetch().
  • Pagination: Use coll.docs.list_pages(size=10, ref=...) to iterate pages of up to size documents, or coll.docs.iter_all(page_size=100, ref=...) to iterate over all documents. Async variants preserve the same ref on every request. Internal document requests support up to 1000 items; the default page size remains 100. See read limits and stored text for Query/Fetch/List limits, unindexed native inputs, and result downloads.
  • Advanced options: Pass options=RequestOptions(timeout_ms=..., http_headers=...) to any docs or query call; import with from lambdadb import RequestOptions. For delete by filter, prefer query_filter=... over filter_=.... Response types such as ListDocsResponse, QueryCollectionResponse, and FetchDocsResponse are also exported from lambdadb for type hints.
Available methods

Collections

  • list - List all collections in an existing project.
  • create - Create a collection.
  • delete - Delete an existing collection.
  • get - Get metadata of an existing collection.
  • update - Configure a collection.
  • query - Search a collection with a query and return the most similar documents.

Collections.Docs

  • list_docs - List documents in a collection.
  • list_pages - Iterate pages of up to size documents each.
  • iter_all - Iterate over all documents (handles pagination).
  • upsert - Upsert documents into a collection. Note that the maximum supported payload size is 6MB.
  • get_bulk_upsert - Request required info to upload documents.
  • bulk_upsert - Bulk upsert documents into a collection. Note that the maximum supported object size is 200MB.
  • bulk_upsert_docs - One-step bulk upsert: upload a list of documents without handling presigned URL or S3 yourself.
  • update - Update documents in a collection. Note that the maximum supported payload size is 6MB.
  • delete - Delete documents by document IDs or query filter from a collection.
  • fetch - Lookup and return documents by document IDs from a collection.

Data Versioning

  • Collection-scoped Branch, Tag, and Alias lifecycle operations.
  • Ref-scoped Query, Fetch, List, and pagination helpers.
  • Branch-scoped document writes and signed bulk uploads.

Retries

Some of the endpoints in this SDK support retries. If you use the SDK without any configuration, it will fall back to the default retry strategy provided by the API. However, the default retry strategy can be overridden on a per-operation basis, or across the entire SDK.

To change the default retry strategy for a single API call, simply provide a RetryConfig object to the call:

from lambdadb import LambdaDB
from lambdadb.utils import BackoffStrategy, RetryConfig

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
) as client:
    res = client.collections.list(
        retries=RetryConfig("backoff", BackoffStrategy(1, 50, 1.1, 100), False)
    )
    print(res)

If you'd like to override the default retry strategy for all operations that support retries, you can use the retry_config optional parameter when initializing the SDK:

from lambdadb import LambdaDB
from lambdadb.utils import BackoffStrategy, RetryConfig

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
    retry_config=RetryConfig("backoff", BackoffStrategy(1, 50, 1.1, 100), False),
) as client:
    res = client.collections.list()
    print(res)

Error Handling

LambdaDBError is the base class for all HTTP error responses. It has the following properties:

Property Type Description
err.message str Error message
err.status_code int HTTP response status code eg 404
err.headers httpx.Headers HTTP response headers
err.body str HTTP body. Can be empty string if no body is returned.
err.raw_response httpx.Response Raw HTTP response
err.data Optional. Some errors may contain structured data. See Error Classes.

Example

from lambdadb import LambdaDB, errors

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
) as client:
    res = None
    try:
        res = client.collections.list()
        print(res)
    except errors.LambdaDBError as e:
        # The base class for HTTP error responses
        print(e.message)
        print(e.status_code)
        print(e.body)
        print(e.headers)
        print(e.raw_response)

        # Depending on the method different errors may be thrown
        if isinstance(e, errors.UnauthenticatedError):
            print(e.data.message)  # Optional[str]

Error Classes

Primary errors:

Less common errors

Network errors:

Inherit from LambdaDBError:

* Check the method documentation to see if the error is applicable.

Server Selection

Use base_url and project_name so the client uses the REST path {base_url}/projects/{project_name} (e.g. https://api.lambdadb.ai/projects/playground).

Parameter Default Description
base_url: str "https://api.lambdadb.ai" API base URL.
project_name: str "playground" Project name (path segment).
from lambdadb import LambdaDB

with LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
) as client:
    res = client.collections.list()
    print(res)

Legacy (deprecated)

server_url and project_host are deprecated and will be removed in the next major version. Prefer base_url and project_name above.

Variable Parameter Default (legacy) Description
projectHost project_host: str "api.lambdadb.com/projects/default" The project URL of the API
# Deprecated: use base_url + project_name instead
with LambdaDB(
    project_host="api.lambdadb.com/projects/default",
    project_api_key="<YOUR_PROJECT_API_KEY>",
) as lambda_db:
    res = lambda_db.collections.list()
# Deprecated: use base_url + project_name instead
with LambdaDB(
    server_url="https://api.lambdadb.com/projects/default",
    project_api_key="<YOUR_PROJECT_API_KEY>",
) as lambda_db:
    res = lambda_db.collections.list()

Custom HTTP Client

The Python SDK makes API calls using the httpx HTTP library. In order to provide a convenient way to configure timeouts, cookies, proxies, custom headers, and other low-level configuration, you can initialize the SDK client with your own HTTP client instance. Depending on whether you are using the sync or async version of the SDK, you can pass an instance of HttpClient or AsyncHttpClient respectively, which are Protocol's ensuring that the client has the necessary methods to make API calls. This allows you to wrap the client with your own custom logic, such as adding custom headers, logging, or error handling, or you can just pass an instance of httpx.Client or httpx.AsyncClient directly.

For example, you could specify a header for every request that this sdk makes as follows:

from lambdadb import LambdaDB
import httpx

http_client = httpx.Client(headers={"x-custom-header": "someValue"})
client = LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
    client=http_client,
)

or you could wrap the client with your own custom logic:

from lambdadb import LambdaDB
from lambdadb.httpclient import AsyncHttpClient
import httpx

class CustomClient(AsyncHttpClient):
    client: AsyncHttpClient

    def __init__(self, client: AsyncHttpClient):
        self.client = client

    async def send(
        self,
        request: httpx.Request,
        *,
        stream: bool = False,
        auth: Union[
            httpx._types.AuthTypes, httpx._client.UseClientDefault, None
        ] = httpx.USE_CLIENT_DEFAULT,
        follow_redirects: Union[
            bool, httpx._client.UseClientDefault
        ] = httpx.USE_CLIENT_DEFAULT,
    ) -> httpx.Response:
        request.headers["Client-Level-Header"] = "added by client"

        return await self.client.send(
            request, stream=stream, auth=auth, follow_redirects=follow_redirects
        )

    def build_request(
        self,
        method: str,
        url: httpx._types.URLTypes,
        *,
        content: Optional[httpx._types.RequestContent] = None,
        data: Optional[httpx._types.RequestData] = None,
        files: Optional[httpx._types.RequestFiles] = None,
        json: Optional[Any] = None,
        params: Optional[httpx._types.QueryParamTypes] = None,
        headers: Optional[httpx._types.HeaderTypes] = None,
        cookies: Optional[httpx._types.CookieTypes] = None,
        timeout: Union[
            httpx._types.TimeoutTypes, httpx._client.UseClientDefault
        ] = httpx.USE_CLIENT_DEFAULT,
        extensions: Optional[httpx._types.RequestExtensions] = None,
    ) -> httpx.Request:
        return self.client.build_request(
            method,
            url,
            content=content,
            data=data,
            files=files,
            json=json,
            params=params,
            headers=headers,
            cookies=cookies,
            timeout=timeout,
            extensions=extensions,
        )

client = LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
    async_client=CustomClient(httpx.AsyncClient()),
)

Resource Management

The LambdaDB class implements the context manager protocol and registers a finalizer function to close the underlying sync and async HTTPX clients it uses under the hood. This will close HTTP connections, release memory and free up other resources held by the SDK. In short-lived Python programs and notebooks that make a few SDK method calls, resource management may not be a concern. However, in longer-lived programs, it is beneficial to create a single SDK instance via a context manager and reuse it across the application.

from lambdadb import LambdaDB

def main():
    with LambdaDB(
        project_api_key="<YOUR_PROJECT_API_KEY>",
        base_url="https://api.lambdadb.ai",
        project_name="playground",
    ) as client:
        # Rest of application here...
        coll = client.collection("my_collection")
        coll.docs.list()

# Or when using async:
async def amain():
    async with LambdaDB(
        project_api_key="<YOUR_PROJECT_API_KEY>",
        base_url="https://api.lambdadb.ai",
        project_name="playground",
    ) as client:
        # Rest of application here...

Debugging

You can setup your SDK to emit debug logs for SDK requests and responses.

You can pass your own logger class directly into your SDK.

from lambdadb import LambdaDB
import logging

logging.basicConfig(level=logging.DEBUG)
client = LambdaDB(
    project_api_key="<YOUR_PROJECT_API_KEY>",
    base_url="https://api.lambdadb.ai",
    project_name="playground",
    debug_logger=logging.getLogger("lambdadb"),
)

You can also enable a default debug logger by setting an environment variable LAMBDADB_DEBUG to true or 1 (case-insensitive). Any other value, including false and 0, leaves it disabled.

Development

Maturity

This SDK is in beta, and there may be breaking changes between versions without a major version update. Therefore, we recommend pinning usage to a specific package version. This way, you can install the same version each time without breaking changes unless you are intentionally looking for the latest version.

Contributions

We welcome contributions. This SDK is maintained manually; open pull requests against main as described in CONTRIBUTING.md, or open an issue.

Metadata

Release files for lambdadb 0.14.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for lambdadb 0.14.1
File Size Uploaded Provenance Actions
lambdadb-0.14.1.tar.gz 92.4 kB With provenance Details for lambdadb-0.14.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for lambdadb 0.14.1
File Interpreter ABI Platform Actions
lambdadb-0.14.1-py3-none-any.whl Python 3 none any Details for lambdadb-0.14.1-py3-none-any.whl

Total release size: 213.3 kB

Release files / lambdadb-0.14.1.tar.gz

Download URL lambdadb-0.14.1.tar.gz
Size 92.4 kB
Tags Source
SHA-256 checksum
How to use checksums
0907919bb39dcbaaefb349c8b20d1b171b22718084fcf4b27332aa2839499859
BLAKE2b-256 checksum
How to use checksums
eb6c831f45d47b70a2b8d581587e31fca877ad42a490f2aa9a7a99289aaf3370
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 30, 2026.

Transparency log

Release files / lambdadb-0.14.1-py3-none-any.whl

Download URL lambdadb-0.14.1-py3-none-any.whl
Size 121.0 kB
Tags Python 3
SHA-256 checksum
How to use checksums
0f1c90c84b276ce0e12136ef7fe07cb8122ba25a6b613a6e0cfd873093b548a7
BLAKE2b-256 checksum
How to use checksums
f1819dc3d10aa792512723ca288eab00d546bc9d6b2d5beaea067d0b359f694a
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 30, 2026.

Transparency log

Security for lambdadb 0.14.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

Both files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for lambdadb-0.14.1.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 30, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.14.1 This release

2 release files Full provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page