Switch HF default model to zephyr-7b-beta (ungated, no license needed)

Qwen/Qwen2.5-7B-Instruct had no visible license gate for the user.
HuggingFaceH4/zephyr-7b-beta is fully open with no access restrictions.
HF_MODEL env var can still override to any other model.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

src/soci/api/routes.py

@@ -278,7 +278,7 @@ async def get_llm_providers():
         providers.append({"id": "groq",    "label": "Groq (Llama 8B)",    "icon": "⚡"})
     if os.environ.get("GEMINI_API_KEY"):
         providers.append({"id": "gemini",  "label": "Gemini 2.0 Flash",   "icon": "✦"})
-    providers.append(    {"id": "hf",      "label": "HF Qwen 2.5 7B",     "icon": "🤗"})
+    providers.append(    {"id": "hf",      "label": "HF Zephyr 7B",        "icon": "🤗"})
     providers.append(    {"id": "ollama",  "label": "Ollama (local)",      "icon": "🦙"})
     return {"current": current, "providers": providers}
 

src/soci/engine/llm.py

@@ -42,6 +42,7 @@ MODEL_GEMINI_FLASH = "gemini-2.0-flash"
 MODEL_GEMINI_PRO = "gemini-1.5-pro"
 
 # Hugging Face Serverless Inference model IDs (free, no credit card)
+MODEL_HF_ZEPHYR = "HuggingFaceH4/zephyr-7b-beta"    # default — ungated, no license needed
 MODEL_HF_LLAMA = "meta-llama/Llama-3.2-3B-Instruct"
 MODEL_HF_QWEN = "Qwen/Qwen2.5-7B-Instruct"
 MODEL_HF_MISTRAL = "mistralai/Mistral-7B-Instruct-v0.3"
@@ -821,7 +822,7 @@ class HFInferenceClient:
     def __init__(
         self,
         api_key: Optional[str] = None,
-        default_model: str = MODEL_HF_QWEN,
+        default_model: str = MODEL_HF_ZEPHYR,
         max_retries: int = 3,
     ) -> None:
         self.api_key = api_key or os.environ.get("HF_TOKEN", "")
@@ -1017,7 +1018,7 @@ def create_llm_client(
         default_model = model or os.environ.get("GEMINI_MODEL", MODEL_GEMINI_FLASH)
         return GeminiClient(default_model=default_model)
     elif provider == PROVIDER_HF:
-        default_model = model or os.environ.get("HF_MODEL", MODEL_HF_QWEN)
+        default_model = model or os.environ.get("HF_MODEL", MODEL_HF_ZEPHYR)
         return HFInferenceClient(default_model=default_model)
     elif provider == PROVIDER_OLLAMA:
         default_model = model or os.environ.get("OLLAMA_MODEL", MODEL_LLAMA)