| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| latest.json | 2026-08-04 | 2.4 kB | |
| Atomic.Chat_2.0.6_x64-setup.exe.sig | 2026-08-04 | 424 Bytes | |
| Atomic.Chat_2.0.6_x64_en-US.msi | 2026-08-04 | 112.3 MB | |
| Atomic.Chat_2.0.6_x64-setup.exe | 2026-08-04 | 75.6 MB | |
| Atomic.Chat.app.tar.gz | 2026-08-04 | 341.3 MB | |
| Atomic.Chat_2.0.6_universal.dmg | 2026-08-04 | 338.4 MB | |
| atomic-chat-mac-universal-2.0.6.zip | 2026-08-04 | 337.6 MB | |
| Atomic.Chat_2.0.6_amd64.AppImage.sig | 2026-08-04 | 424 Bytes | |
| Atomic.Chat_2.0.6_amd64.AppImage | 2026-08-04 | 228.1 MB | |
| Atomic Chat v2.0.6 source code.tar.gz | 2026-08-04 | 235.8 MB | |
| Atomic Chat v2.0.6 source code.zip | 2026-08-04 | 237.1 MB | |
| README.md | 2026-08-04 | 859 Bytes | |
| Totals: 12 Items | 1.9 GB | 0 | |
π New Features
New models on llama.cpp TurboQuant β DeepSeek V4 Flash and Kimi K3
- DeepSeek V4 Flash 0731 runs
- Kimi K3 fully supported, vision included
- AMD on Linux x64 now covers RDNA2βRDNA4 and CDNA
- Synced with the latest upstream llama.cpp
π§ Improvements & Fixes
- Downloads no longer fail when the pre-check does
- Clearer model-load errors, and recoverable ones now retry
- Windows: models in folders with non-Latin names now load
- MLX models with legacy or misnamed shards now load
- Favorites no longer list the same model twice
- Proper icon for the Codex integration
- Smaller UI and stability cleanups
π Contributors
Thanks to @Vect0rM, @worthant, @yl-atomic for their contributions to this release!