v1.9.0

Changelog

Features 🪄

• #571 add discovery and recommendations api support for sparse vectors in local mode
• #594 restrict access to the methods on a closed instance in local mode
• #591 add a possibility to use auth token providers
• #608 introduce support for uint8 vectors
• #603 allow having nan values in payload in local mode

---

Fixes 🪛

• #559 - fix update_vectors in local mode with sparse vectors
• #562 - fix handling of wrong types in search in local mode
• #560 - check vectors for nan values
• #569 - fix grpc sparse vector conversion
• #575 #576 #580 - fix datetime and date comparison
• #581 #583 fix sparse vectors async api
• #584 fix data-parallelism with sparse models
• #587 fix read consistency with grpc in batch search
• #588 fix migrate vector count
• #605 fix grpc field schema type conversion for datetime index

---

Deprecations ⏳

• #608 deprecate recreate collection method
• #608 remove vectors count

---

Thanks everyone who contributed to the release @coszio @yasyf @Apmats @skvark @generall @joein @agourlay

v1.8.2

Changelog

Fixes 🪛

• remove redundant pytest import which prevented qdrant-client import without pytest being installed by @joein

v1.8.1

Changelog

Features 🪄

• update fastembed integration, incorporate hybrid search into add method #553 by @generall @joein

---

Fixes 🪛

• fix excess upsert requests in upload_collection and upload_points #531 by @almostimplemented
• allow set payload with a nested key #536 by @joein
• add missing datetime formats to local mode #537 by @joein
• fix congruence for naive and aware datetime in local mode #538 by @joein
• propagate timeouts from methods to httpx #534 by @joein

---

Thanks everyone who contributed to the release @almostimplemented @generall @joein

v1.8.0

Changelog

---

Features 🪄

• add order_by functionality to scroll #491 by @coszio
• add grpc compression options to QdrantClient #480 by @geetu040
• add datetime ranges support #517 by @coszio
• add min should clause #519 by @joein
• add collection exists api #518 by @joein
• update fastembed integration #493 #522 #524 @NirantK @joein @Anush008

---

Fixes 🪛

• fix potential edge case in scoring in context search #474 by @coszio
• close lock file explicitly in local mode #505 #521 by @FranckZibi @joein
• fix async client generator messing up imports #480 by @joein

---

Deprecations ⏳

• remove deprecated field conversions in grpc by @coszio

---

Thanks everyone who contributed to the release @coszio @geetu040 @FranckZibi @NirantK @Anush008 @joein

v1.7.3

• fix search_batch with sparse vectors in grpc #484 by @agourlay
• align close method signature between QdrantClient and QdrantRemote #477 by @joein

v1.7.2

Bugfix release 💊

• fix parallel and bugs with implicit ids in upload_collection, upload_records, upload_points #460
• fix setting grpc_options with https in sync mode, and with http+api-key in async mode #467
• fix extend json validation in local mode to fix backward compatibility issue with types like datetime and uuid #462

v1.7.1

Changelog

---

Features 🪄

• exposed gRPC options #401
• support setting custom shard key in upload_collection and upload_points #396
• unlocked python dependency upper cup #419
• unlocked urllib3 v2 #418
• ease access to fastembed models #412
• extend access for fastembed configuration #416

---

Fixes 🪛

• replaced httpx data with content #426
• forbid having NaN in payload in local mode #397
• fix python warning which was sometimes emitted on garbage collection in local mode #398 #421
• fix division by zero error in cosine similarity in local mode #425
• add wait parameter to upload_records in migrate to avoid false-positive asserts #428
• fix operations with sparse vectors with unsorted indices in local mode #442
• restrict inserting points with non-configured vectors in local mode #432
• fix reading collections with unnamed dense and named sparse vectors in local mode #433
• forbid to migrate from collections with custom shard keys #447
• align timeout param types in QdrantClient and its methods

Deprecation ⏳

• upload_records has been deprecated in favour of upload_points #447

Thanks to the devs who contributed to this release

@generall @shivas1516 @praveen-palanisamy @joein

v1.7.0

Changelog

---

Features

• Add Discovery API (docs) - #368 by @coszio
  • Discovery search: Uses a target and context pairs to limit the search space during the search
  • Context search: Using only the context pairs, get the points that live in the best zone
• Add Sparse vectors support (docs) - #378 by @agourlay
  • Combine the dense and sparse vectors to build hybrid search
• Add Manhattan distance support - #391 by @generall
• Add custom shard key API (docs) - #391 by @generall
  • Configure how to distribute points among shards

---

Fixes

• #337 - Add wait parameter to snapshots API by @joein

v1.6.2

Changelog

• Missing fix form previous release: #338 - add missing parameters in recommend batch API

v1.6.1

Changelog

Features

• Proper support for Async client - #319

  • Now all methods of regular qdrant client are available in async version.
  • All method names, parameters and return types are preserved.
  • Both gRPC and REST version are available.

• Improvements of fastembed integration:

  • Support the latest v0.1.x version of fastembed: https://github.com/qdrant/fastembed/releases/tag/0.1.0
  • Support for encoding of iterators in add API, useful for lazy reading of large dataset
  • Support for data-level parallelism of the encoding operations

Fixes

• #337 - fix: convert score to float in local mode for pydantic

---

Async Qdrant client usage example:

from qdrant_client import AsyncQdrantClient, models
import numpy as np
import asyncio

async def main():
    client = AsyncQdrantClient(url="http://localhost:6333")

    res = await client.search(
        collection_name="my_collection",
        query_vector=np.random.rand(10).tolist(),  # type: ignore
        limit=10,
    )

    print(res)

asyncio.run(main())