Atomic Chat v1.1.53

🚀 New Features

Bring your own API keys — connect to 10+ top model providers including OpenRouter, Anthropic, HuggingFace, MiniMax and more. Works in chat and via the local API server
2x faster local models with MLX on Apple Silicon
Concurrent Mode — run up to 16 Gemma 4 instances in parallel on your local API server
Reasoning control — toggle thinking on or off per chat on supported models
Notifications — get alerted when a reply finishes while the window is in the background
Menu bar status — see your server, active model and RAM usage at a glance