SalesforceAIResearch/xLAM
A family of Large Action Models designed to enable AI agents to plan, reason, and execute multi-step actions across diverse tasks.
Not currently ranked — collecting fresh signals.
star history
xLAM provides a series of Large Action Models purpose-built for AI agent systems. The repository includes training frameworks like ActionStudio and LATTE for developing action models, evaluation benchmarks to measure agent performance, and supporting infrastructure for deploying these models in agent pipelines.
Frequently asked
- What is SalesforceAIResearch/xLAM?
- A family of Large Action Models designed to enable AI agents to plan, reason, and execute multi-step actions across diverse tasks.
- Is xLAM open source?
- Yes — SalesforceAIResearch/xLAM is open source, released under the Apache-2.0 license.
- What language is xLAM written in?
- SalesforceAIResearch/xLAM is primarily written in Python.
- How popular is xLAM?
- SalesforceAIResearch/xLAM has 634 stars on GitHub.
- Where can I find xLAM?
- SalesforceAIResearch/xLAM is on GitHub at https://github.com/SalesforceAIResearch/xLAM.