Upload README.md with huggingface_hub

README.md

@@ -1,13 +1,17 @@
 ---
-title: Supra Router 51m
-emoji: ⚡
-colorFrom: purple
-colorTo: blue
+title: Supra-Router-51M
+emoji: 🛤️
+colorFrom: blue
+colorTo: indigo
 sdk: gradio
 sdk_version: 6.19.0
-python_version: '3.13'
 app_file: app.py
 pinned: false
+short_description: "51M prompt router: edge SLM vs cloud frontier"
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Supra-Router-51M
+
+A Gradio demo for [SupraLabs/Supra-Router-51M](https://huggingface.co/SupraLabs/Supra-Router-51M), an ultra-lightweight (51.7M parameter) multi-task infrastructure routing model. Given a user prompt, it predicts whether the request should be handled locally by an edge SLM or offloaded to a cloud frontier model, along with a structured analysis of the prompt's domain, complexity, math/code flags, and a justification.
+
+The model wraps the query as `Task: {prompt}\nAnalysis: ` and performs greedy decoding to emit a deterministic, pipe-separated decision string. This Space runs entirely on CPU — the model is small enough for fast local inference.