polish: add tags + model refs + thumbnail for HF discoverability

README.md

@@ -9,8 +9,31 @@ app_file: app.py
 pinned: false
 license: mit
 short_description: Chat with DeepSeek R1 / V3 / Qwen 3.5 via QuickSilver Pro
+tags:
+  - chatbot
+  - llm
+  - deepseek
+  - deepseek-r1
+  - deepseek-v3
+  - qwen
+  - qwen3
+  - openai-compatible
+  - streaming
+  - gradio
+  - inference
+  - api
+models:
+  - deepseek-ai/DeepSeek-V3
+  - deepseek-ai/DeepSeek-R1
+  - Qwen/Qwen2.5-32B
+thumbnail: https://quicksilverpro.io/og.png
 ---
 
+<!-- Thumbnail uses the existing site OG image (1200×630 indigo-purple brand
+card). HF will automatically generate one from colorFrom/colorTo if this URL
+is unreachable — safe fallback. -->
+
+
 # QuickSilver Pro Chat
 
 Try **DeepSeek V3**, **DeepSeek R1**, and **Qwen 3.5-35B-A3B** via an OpenAI-compatible endpoint — no signup required.