tomoul doctor
Diagnose your local install. Prints the engine version, GPU detection, available quantizations, and a sample inference test.
Usage
tomoul doctorRun this before opening a bug report — paste the output into the issue and we'll know exactly what your machine looks like.
Sample output
tomoul 0.4.2 (engine 0.4.2-zig)
OS: darwin arm64
Device: Apple M2 Max (38 GPU cores, Metal 3)
Memory: 64 GB
Quant: fp16 ✓ int8 ✓ int4 ✗ (int4 needs CUDA)
Cache: /Users/you/.cache/tomoul (4 models, 12.4 GB)
Network: api.tomoul.ai reachable (37ms)
Smoke test: bge-m3 → ✓ (94ms cold, 8ms warm)Last updated 13 May 2026Edit this page on GitHub