- Glean Assistant now supports LLM choice so users can pick the best-suited model for each task (creative, coding, or general-purpose), letting AI adapt to their work instead of forcing workflows around a single model.
- This flexibility improves outcomes by making models from multiple providers instantly accessible in one place, enabling practical, side-by-side evaluation on real workloads and continuously updating “best model” defaults as models evolve.
- Enterprises retain strong governance and compliance through an admin console that controls which model families are available, how they’re used, and who can access them, while avoiding vendor lock-in and maintaining transparency into model behavior and tradeoffs.
AI should flex to fit your work so that you can be an expert in any task and have an agentic coworker by your side. That belief is at the core of how we built Glean Assistant, and it’s exactly why we’re introducing LLM choice. LLM choice gives teams the ability to select the AI model that best fits the task they are trying to accomplish, providing them the flexibility they need to adapt in an environment where new models are regularly introduced.
Different work requires different strengths
Each new model launches with different strengths, tradeoffs, and cost profiles. LLM choice makes it easy to evaluate alternatives and adapt as needs and options change.
Glean now makes it easy to choose the model that best supports the task you’re trying to accomplish. LLMs are accessible as soon as they’re available (across providers), without you needing to switch tools or sign new contracts.
- Designing new creative assets? Choose a model optimized for writing and image generation like Gemini 3.1 Pro.
- Engineering tasks demand code-specialized models that understand syntax and structure. Switch to a coding-focused model like Claude Sonnet 4.6.
- Doing lots of different activities in a single session? Select a strong all-purpose model like GPT 5.2.
This flexibility isn’t just about having more options. It’s about driving better outcomes. When the AI is well-matched to the work, answers are more accurate, insights come faster, and users have greater confidence in the results. Instead of adapting workflows to fit the limitations of a single model, teams can let AI adapt to them.
Flexibility paired with enterprise oversight and compliance
We built LLM choice with both end users and enterprise administrators in mind. For any task, users can experiment with different models on real workloads, directly comparing outputs and performance without leaving Glean. This makes model evaluation practical and grounded in real work. Behind the scenes, Glean continuously tracks live usage patterns and evaluations to keep recommended “best model” defaults accurate over time. As models evolve, Glean adapts — so customers don’t have to constantly re-evaluate the landscape on their own.
At the same time, administrators maintain full control over model usage and choice. From the admin console, they can curate which model families are available, define usage policies, and govern access across the organization. Teams can access the latest large language models instantly, without waiting for system-wide updates or switching between tools. This enables thoughtful, strategic model governance of the latest models. Teams gain flexibility, while enterprises retain oversight and compliance.
Why it matters
Glean Assistant is your agentic coworker that is deeply proficient at any task. With LLM choice, users can use the right model optimized for any task they want to accomplish. Simultaneously, organizations avoid vendor lock-in, gain transparency into model behavior and tradeoffs, and can innovate with confidence. No matter the work, you always have access to leading innovation.
If you want to learn more about how to start leveraging your own team of experts through Glean Assistant and LLM choice, reach out to our team for a demo and check out our March drop page for our latest feature releases.
