Model Hub
Models that match the work.
Glean’s Model Hub gives teams 40+ open and proprietary LLMs, with auto routing and controls for access, safety, and spend.
5 trillion+Tokens routed each month
81% Lower token costs as compared to Claude Cowork
40+ Open and proprietary models available
A single hub to choose and control models for enterprise work.
Access models without lock in
Choose from leading open and proprietary models without being tied to one provider or missing what’s new.
Control access and spend
Choose which models are available and who can use them, then set usage limits to manage spend as adoption grows.
Secure models wherever they work
AI models are secured against prompt injection and your data is kept safe at runtime, no matter which model you're using, frontier or open.
Make better decisions with full context
Get proactive answers, recommendations, and next steps grounded in your company’s full context, so every decision starts with the right information.
Make better decisions with full context
Get proactive answers, recommendations, and next steps grounded in your company’s full context, so every decision starts with the right information.
Auto routing
Beta
Match the right model to the job.
Auto routing selects from 40+ LLMs based on performance and cost to match the right model to the job, delivering frontier-level intelligence without the frontier cost.
Governance
Model choice on your terms.
AI usage controls let admins turn models on or off, control model access by user or department, and set usage limits to manage spend by user, agent, or department.
AI Gateway
Manage mode governance from one place.
Glean’s AI Gateway lets enterprises manage and control LLM usage and spend across AI applications centrally, extending the model hub’s access and governance beyond Glean.
Glean Universal Model Key
Beta
One secure key to access leading models.
Get ZDR-compliant access to leading open and frontier models across providers. Your data isn’t used for model training, with consistent security across open and proprietary models.
