docs: align Space copy with deployed workflow

README.md

@@ -8,19 +8,21 @@ sdk_version: 5.50.0
 app_file: app.py
 pinned: true
 license: mit
-short_description: "AI Video Clipping: speak to clip, powered by FunASR + LLM"
+short_description: "Chinese sentence-based video clipping with FunASR"
 ---
 
-# FunClip: AI-Powered Video Clipping
+# FunClip: ASR-Guided Video Clipping
 
-Upload a video → FunASR transcribes all speech with timestamps → Select segments by text → Export precise clips automatically.
+Upload a Chinese-language video → FunASR Paraformer transcribes it with sentence-level timestamps → Select one or more sentences → Export the selected segments as a clip.
 
-## Features
-- 🎬 Automatic speech-to-text with word-level timestamps
-- ✂️ Click on any sentence to create a clip
-- 🤖 LLM-assisted smart clipping (find highlights automatically)
-- 🌍 Multi-language support (Chinese, English, Japanese, Korean, etc.)
+## What This Space Does
+- Transcribes Chinese speech with FunASR Paraformer on CPU
+- Returns sentence-level timestamps for each recognized segment
+- Lets you select and concatenate one or more sentences
+- Exports the result as an MP4 with FFmpeg
+
+Short videos provide the best interactive experience on the shared CPU runtime. For multilingual ASR and LLM-assisted smart clipping, install the full [FunClip](https://github.com/modelscope/FunClip) app locally.
 
 ## Links
-- **GitHub**: [FunClip](https://github.com/modelscope/FunClip) (⭐ 5.6k+)
+- **GitHub**: [Star the FunClip repository](https://github.com/modelscope/FunClip)
 - **ASR Engine**: [FunASR](https://github.com/modelscope/FunASR) | [Fun-ASR](https://github.com/FunAudioLLM/Fun-ASR)

app.py

@@ -131,12 +131,12 @@ def clip_video(video_path, sentences_json, selected_indices):
 description_html = """
 <div style="text-align: center; max-width: 850px; margin: 0 auto;">
     <h1 style="font-size: 2.2em; margin-bottom: 0.1em;">✂️ FunClip</h1>
-    <p style="font-size: 1.3em; color: #444;">AI Video Clipping — Speak to Clip</p>
+    <p style="font-size: 1.3em; color: #444;">Chinese Video Clipping with FunASR</p>
     <p style="font-size: 1em; color: #666;">
-        Upload a video → Auto-transcribe with timestamps → Select text segments → Export precise clips
+        Upload a Chinese-language video → Transcribe with sentence-level timestamps → Select segments → Export a clip
     </p>
     <p style="font-size: 0.9em; margin-top: 0.8em;">
-        <a href="https://github.com/modelscope/FunClip" target="_blank">⭐ GitHub (5.6k+ stars)</a> ·
+        <a href="https://github.com/modelscope/FunClip" target="_blank">⭐ Star FunClip on GitHub</a> ·
         <a href="https://github.com/modelscope/FunASR" target="_blank">🛠️ FunASR</a> ·
         <a href="https://github.com/FunAudioLLM/Fun-ASR" target="_blank">🚀 Fun-ASR</a>
     </p>
@@ -146,11 +146,11 @@ description_html = """
 how_it_works = """
 ### How It Works
 1. **Upload** a video (any format with audio)
-2. **Transcribe** — FunASR extracts speech with precise timestamps
+2. **Transcribe** — Chinese Paraformer ASR adds sentence-level timestamps
 3. **Select** the sentences you want to keep (by index)
 4. **Clip** — FFmpeg cuts and concatenates the selected segments
 
-For the full experience with LLM-assisted smart clipping, install [FunClip](https://github.com/modelscope/FunClip) locally.
+For multilingual ASR and LLM-assisted smart clipping, install the full [FunClip](https://github.com/modelscope/FunClip) app locally.
 """