Buddy's brain: Gemma 4 26B-A4B
The on-device model used by the iOS app Buddy. The app downloads these files and runs them on the phone.
| File | What |
|---|---|
buddy-gemma4.resident.gguf |
The always-used part (from Unsloth's UD-IQ2_M, 5-bit) |
buddy-gemma4.experts |
The experts, read from flash (from Unsloth's UD-IQ3_S) |
buddy-gemma4.json |
Where each expert sits in the experts file |
buddy-gemma4-26b-eyes-Q8_0.gguf |
Vision projector (ggml-org's mmproj Q8_0) |
Made from Google's Gemma 4 26B-A4B (Apache 2.0) via Unsloth's and ggml-org's GGUF files, unchanged except for the split. Licensed Apache 2.0.
- Downloads last month
- 337
Hardware compatibility
Log In to add your hardware
8-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support