Auto Repair & Service Ops
Books service, quotes repairs, and answers shop questions for auto businesses.
- Base model
- Qwen3.6-35B-A3B
- License
- Apache-2.0
- Version
- v1.0
Run it locally
Free · open weightsApache-2.0 weights on your own box. Your data never leaves the machine, and a lapsed plan never bricks a local model.
Hugging Face
shell
huggingface-cli download flywheel-ai/automotive # serve OpenAI-compatible with vLLM: vllm serve flywheel-ai/automotive --served-model-name automotive
Use the hosted API
OpenAI-compatibleDon’t want to run hardware? Point any OpenAI SDK at Flywheel — change one line (the base_url) and use a fw_live_ key.
curl https://gyld.dev/api/v1/chat/completions \
-H "Authorization: Bearer $FLYWHEEL_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "automotive",
"messages": [
{ "role": "user", "content": "Quote a brake job and book the next open service slot." }
]
}'from openai import OpenAI
client = OpenAI(
base_url="https://gyld.dev/api/v1", # the one line you change
api_key="fw_live_...", # mint on the Account page
)
resp = client.chat.completions.create(
model="automotive",
messages=[{"role": "user", "content": "Quote a brake job and book the next open service slot."}],
)
print(resp.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://gyld.dev/api/v1", // the one line you change
apiKey: "fw_live_...", // mint on the Account page
});
const resp = await client.chat.completions.create({
model: "automotive",
messages: [{ role: "user", content: "Quote a brake job and book the next open service slot." }],
});
console.log(resp.choices[0].message.content);Builds & sizes
| Format | Size | Best for |
|---|---|---|
| Q4_K_M GGUF | 20GB | Laptops / Ollama / llama.cpp |
| bf16 safetensors | ~65GB | GPU serving (vLLM, TGI) & fine-tuning |
Limitations
Output is decision support, not professional advice. Verify anything consequential with a licensed professional in your field.