== 1. listener on port 8000
  ollama *:8000
  PASS  ollama listens on all interfaces, so VMs can reach it
== 2. HTTP on http://127.0.0.1:8000
  PASS  Ollama 0.33.3 answers on http://127.0.0.1:8000
== 3. VM-facing address 192.168.64.1
  PASS  Ollama answers on http://192.168.64.1:8000 (bridge100), the address VMs use
== 4. launchd config
  info  OLLAMA_HOST in plist: 0.0.0.0:8000
== 5. models
  PASS  gemma4:e2b is installed
== 6. native API
  23 tokens @ 43.6 tok/s, load 5.7 s
  PASS  gemma4:e2b answered through /api/generate
== 7. streaming
  PASS  streaming delivers incremental chunks
== 8. OpenAI-compatible API
  PASS  /v1/chat/completions returned text
== 9. GPU
  PASS  gemma4:e2b loaded 100% on the GPU (1.6 GB)
== 10. second model: gemma4:e2b-it-qat
  33 tokens @ 43.3 tok/s, load 7.5 s
  PASS  gemma4:e2b-it-qat answered through /api/generate
  PASS  gemma4:e2b-it-qat loaded 100% on the GPU (3.3 GB)

10 passed, 0 failed
Ollama works on the Mac. Next: bin/test-vm-ollama