Use fabryka in Hermes
Hermes reads its model from ~/.hermes/config.yaml — not env vars. One block is the whole integration.
1 · Get a key
Grab a free $100 key at router.fabryka.ai (email, no card).
2 · Edit ~/.hermes/config.yaml
yaml
# ~/.hermes/config.yaml
model:
default: qwen3.6-35b-a3b
provider: custom:fabryka
base_url: https://router.fabryka.ai/v1
api_mode: chat_completions
api_key: sk-fab-YOUR_KEY # from step 13 · Restart Hermes
bash
sudo systemctl restart hermes-sdk-server # or however you run hermes4 · Verify
bash
curl http://127.0.0.1:8800/v1/chat/completions \
-H "Authorization: Bearer $YOUR_LOCAL_HERMES_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"qwen3.6-35b-a3b","messages":[{"role":"user","content":"reply: online"}]}'Notes
• Only the LLM api_key changes to your sk-fab-…. The key your app uses to reach the local Hermes SDK is separate and stays the same.
• qwen3.6-35b-a3b reasons by default — for fast tool steps send chat_template_kwargs: {enable_thinking:false}.
• Single-GPU backend → keep concurrency at 1.