---
title: "Version 3.4"
slug: "ml-version-34"
updated: 2025-10-20T10:24:51Z
published: 2025-10-20T10:24:51Z
canonical: "docs.knovvu.com/ml-version-34"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.knovvu.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Version 3.4

## Version 3.4
### 24 June, 2025

:::(Error) (Important Note)
Starting from v3.4, the **question-answering-v21** module will be using **Qdrant Vector Database** to store knowledge base embeddings, replacing the previous in-memory storage for improved scalability and performance. As a result, question-answering-v21 must be used with Knovvu VA & Knovvu ML v3.4 and it is **not backward compatible** with earlier versions.
:::

:::(Error) (Important Note)
Starting from v3.4, the **ai-agent-v3** module must be used with Knovvu VA & Knovvu ML v3.4 and it is **not backward compatible** with earlier versions.
:::

:::(Error) (Important Note)
Starting from v3.4, the **text-classification-mlp-v28** module must be used with Knovvu VA & Knovvu ML v3.4 and it is **not backward compatible** with earlier versions.
:::

:::(Info) (AI Module Note)
Please use the following stable AI modules for the v3.4 update! 

* ai-agent (v3) -- Optional service, requires Qdrant Vector Database
* question-answering (v21) -- Requires Qdrant Vector Database
* generative-ai (v4)
* sentence-embedding (v23)
* text-normalization (v30)
* text-classification-mlp (v28)
* text-sentiment (v18)
* text-clustering (v28)
* text-language (v13)
* text-summarization (v10) -- Requires Vllm for operation, cannot function independently.

:::

### Improvements

- **Text Clustering with LLM Providers:**
Customers can now use their own API keys to choose from a variety of LLM providers—including OpenAI, Azure, Bedrock, Gemini, Together AI, Groq, and SESTEK for clustering tasks. This flexibility empowers organizations to optimize for cost, performance, and data privacy based on their specific needs. 

**New Version of AI Modules:**
* ai-agent (v3) has been tested and released.
* question-answering (v20) has been tested and released.
* generative-ai (v4) has been tested and released.
* sentence-embedding (v23) has been tested and released.
* text-normalization (v30) has been tested and released.
* text-classification-mlp (v28) has been tested and released.
* text-clustering (v28) has been tested and released.
* text-language (v13) has been tested and released.
* text-summarization (v10) has been tested and released.