We’re releasing a new version of QAnswer with a series of updates focused on model flexibility, multi-user collaboration, image understanding, and improved report generation. These new features come in response to real-world needs expressed by our users—particularly teams working in document-heavy environments where speed, clarity, and adaptability are critical.

This article walks through each of the new features and improvements now available across both SaaS and on-premise deployments.

Model Selection: Switch Between Private and Public LLMs

One of the most requested features is now available: selectable language models. You can now choose, on a per-assistant basis, between

• QAnswer LLM: our proprietary model, optimised for information retrieval, semantic understanding, and deployment in private     or regulated environments.
• Public LLMs: including ChatGPT and Mistral, with others to follow.

This allows organisations to balance performance, compliance, and cost depending on their context. For instance, internal users working with sensitive data can default to QAnswer LLM, while sandbox or exploratory projects might benefit from public LLMs’ wider scope. Switching is seamless within the admin panel, giving more control without complexity.

Image Support: Let the Assistant “See”

QAnswer now supports image-based input, expanding beyond text to enable richer document interpretation.

In practical terms, this means the assistant can process a chart, scan, photo, or infographic, extract relevant content, and answer follow-up questions just like it would with paragraphs of text. This is particularly useful when working with scanned contracts, regulatory filings, or documents that mix text and visuals.

The image is processed in context, so you can reference it naturally in your prompts without needing extra formatting.

Collaborative Chat Sessions

A new collaborative mode is now available in QAnswer chats. Multiple users can now participate simultaneously in the same conversation, enabling real-time input from different individuals within a shared chat session.

This functionality allows teams to work together in a single assistant environment.The system identifies each participant, maintains context across inputs, and ensures continuity in the conversation. It is particularly effective in cross-functional settings—such as when legal, policy, and technical profiles contribute to the same discussion or knowledge analysis.

This update makes QAnswer more suitable for collaborative workflows where input from several team members is required during live interaction with the assistant.

Human Takeover (Upcoming)

We’re preparing a human takeover feature, which allows users interacting with a chatbot to escalate the conversation to a human agent. When activated, a staff member can join the same chat and begin responding manually, using the same interface.

The feature is designed for organisations offering hybrid support: AI handles most queries, but certain interactions require human supervision, validation, or decision-making.

While this functionality is not yet live, the infrastructure is already in place and undergoing internal testing.

Smarter Report Generation: Templates, Structure, and Reusability

Report generation has been a core use case for manyQAnswer users—especially in sectors where summaries, overviews, or document creation workflows are tied to standard formats. We’ve improved this process with three main upgrades:

• Template Creation: Save reusable report structures, reducing friction in repeated tasks.
• Template Import: Bring in formats from Word, Markdown, or HTML and apply them directly.
• Improved Output Control: Generated content now follows formatting and logical flow more closely, reducing the need for post-processing.

Combined, these updates turn QAnswer into a more capable assistant for automated documentation and reporting.

This new QAnswer release reflects our direction: more adaptability, more collaboration, and better integration into existing workflows.Rather than offering generic answers, QAnswer continues to evolve as a context-aware assistant that fits into the way organisations work—whether that’s in a policy office, a research lab, or a support team.

All updates are now available in both SaaS and on-premise environments. We’re continuing to build features aligned with real needs, and as always, feedback from our users is what drives the roadmap forward.

If you're already using QAnswer, these updates are available by default. For new users, feel free to contact us to learn more or request access.