Hermes Agent v0.21.0: Gemma 4 on MLX & GitHub Copilot Integration
Action Required
Users can now leverage the latest Gemma 4 model on Apple Silicon, expanding Hermes' capabilities and potentially improving performance for research and engineering workflows.
AI Impact Summary
Version 0.21.0 of the Hermes Agent introduces significant updates, including support for running Gemma 4 on Apple Silicon via MLX, along with enhancements to the MLX backend like mixed-precision quantization and new op wrappers. The addition of GitHub Copilot CLI integrations and inline config updates streamline workflows. These changes expand Hermes' capabilities for research and engineering tasks, offering improved performance and flexibility for users.
Affected Systems
- Date
- 18 Apr 2026
- Change type
- capability
- Severity
- high