v2.6.13

## v2.6.13

Release date: March 23, 2026


| Milvus Version | Python SDK Version | Node.js SDK Version | Java SDK Version | Go SDK Version |
| -------------- | ------------------ | ------------------- | ---------------- | -------------- |
| 2.6.13         | 2.6.10             | 2.6.11              | 2.6.16           | 2.6.1          |

### Features

#### Gemini embedding model support ([#48223](https://github.com/milvus-io/milvus/pull/48223))

Added Google Gemini as a built-in text embedding function. Users can now use Gemini embedding models directly in Milvus by configuring a Gemini API key, including the recently released [Gemini Embedding 2](https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2/).

For detailed usage, refer to [Google Gemini](google-gemini.md).

### Improvements

- Unified KV path/key conventions across etcd, tikv, and catalog layers with consistent path joining ([#48133](https://github.com/milvus-io/milvus/pull/48133))
- Added query metrics for JSON-related filter expressions to improve observability of JSON field query performance ([#48147](https://github.com/milvus-io/milvus/pull/48147))
- Reduced transient memory allocations in BM25Stats deserialization by eliminating temporary slices ([#48178](https://github.com/milvus-io/milvus/pull/48178))
- Added TruncateCollection method to the Go SDK client for clearing all data in a collection without dropping it ([#48361](https://github.com/milvus-io/milvus/pull/48361))

### Bug fixes

- Fixed an issue where search/query requests with strong consistency timed out during compaction due to tsafe advancement being blocked ([#47987](https://github.com/milvus-io/milvus/pull/47987))
- Fixed an issue where partial upsert with dynamic fields failed with `the length of valid_data of field($meta) is wrong` when the batch contained both existing and new rows ([#48085](https://github.com/milvus-io/milvus/pull/48085))
- Fixed TLS connection failures during internal proxy-to-proxy request forwarding ([#48226](https://github.com/milvus-io/milvus/pull/48226))
- Fixed query failures when using IN combined with != filter expressions that form tautologies ([#48261](https://github.com/milvus-io/milvus/pull/48261))
- Fixed high system load caused by mass concurrent client disconnections during HTTP request handling ([#48270](https://github.com/milvus-io/milvus/pull/48270))
- Fixed an issue where queries could become unavailable during replica scale-up/down due to non-deterministic node-to-replica assignment ([#48277](https://github.com/milvus-io/milvus/pull/48277))
- Fixed HTTP proxy crashes caused by concurrent write race condition in timeout middleware ([#48296](https://github.com/milvus-io/milvus/pull/48296), [#48317](https://github.com/milvus-io/milvus/pull/48317), [#48356](https://github.com/milvus-io/milvus/pull/48356))
- Fixed potential query node crash caused by assertion failure in delete record snapshot handling ([#48302](https://github.com/milvus-io/milvus/pull/48302))
- Fixed storage operation failures when using AK/SK authentication on Aliyun OSS ([#48311](https://github.com/milvus-io/milvus/pull/48311))
- Fixed degraded search performance caused by permanent parameter cache failure leading to goroutine contention on proxy hot paths ([#48313](https://github.com/milvus-io/milvus/pull/48313), [#48326](https://github.com/milvus-io/milvus/pull/48326))
- Fixed QueryCoord deadlock during upgrades when hundreds of channels needed rebalancing by splitting the executor into separate channel and non-channel task pools ([#48351](https://github.com/milvus-io/milvus/pull/48351))
- Fixed an issue where search requests could timeout for 14+ minutes after WAL ownership changes due to unbounded message replay during scanner catchup ([#48391](https://github.com/milvus-io/milvus/pull/48391))