improve style intro & radio button model

app.py

@@ -256,12 +256,12 @@ with gr.Blocks() as demo:
             )
             model_radio = gr.Radio(
                 [*ENHANCEMENT_MODELS],
-                value="FINCH",
+                value="Quail",
                 label="Select Model",
                 info=(
+                    "[QUAIL](https://ai-coustics.com/2025/07/02/meet-quail-the-most-advanced-real-time-speech-enhancement-model/): lightweight, specialized in real-time audio enhancement (no API key needed). "
                     "[FINCH](https://ai-coustics.com/2025/06/12/introducing-finch-2-the-new-aicoustics-model-for-studio-quality-speech/): specialized on voice isolation/removing background noise (API key required). "
                     "[LARK 2](https://ai-coustics.com/2025/07/29/lark-2-next-generation-reconstructive-speech-enhancement/): universal speech enhancement model (API key required)."
-                    "[QUAIL](https://ai-coustics.com/2025/07/02/meet-quail-the-most-advanced-real-time-speech-enhancement-model/): lightweight, specialized in real-time audio enhancement (no API key needed)."
                 ),
             )
             btn = gr.Button("Enhance")

constants.py

@@ -1,7 +1,7 @@
 from typing import Final
 
 
-ENHANCEMENT_MODELS: Final = ["FINCH", "LARK_V2", "QUAIL"]
+ENHANCEMENT_MODELS: Final = [("Quail", "QUAIL"), ("Finch", "FINCH"), ("Lark 2", "LARK_V2")]
 API_V2_URL: Final = "https://api.ai-coustics.io/v2"
 CHUNK_SIZE: Final = 1024
 TIMEOUT_FACTOR_MB: Final = 60

intro.md

@@ -1,14 +1,11 @@
 
-Welcome! This interactive demo lets you denoise and enhance audio files using ai-coustics models.
+Welcome! This interactive demo lets you denoise and enhance audio files using ai-coustics models. Choose from three models, each designed for specific use cases:
 
-You can choose from three models, each designed for different needs:
-- [Quail](https://ai-coustics.com/2025/07/02/meet-quail-the-most-advanced-real-time-speech-enhancement-model/): Our smallest, most lightweight model, ideal for real-time voice isolation. Quail is available for free here and in our [SDK](https://docs.ai-coustics.com/sdk/overview).
-- [Finch](https://ai-coustics.com/2025/06/12/introducing-finch-2-the-new-aicoustics-model-for-studio-quality-speech/): A larger model specialized in voice isolation and background noise removal. Available via our [API](https://docs.ai-coustics.com/api-reference/quickstart) (API key required).
-- [Lark](https://ai-coustics.com/2025/07/29/lark-2-next-generation-reconstructive-speech-enhancement/): Our most advanced, universal speech enhancement model, capable of restoring audio artifacts such as distortion. Also available via our [API](https://docs.ai-coustics.com/api-reference/quickstart) (API key required).
+- [Quail](https://ai-coustics.com/2025/07/02/meet-quail-the-most-advanced-real-time-speech-enhancement-model/): Our smallest, most lightweight model for real-time voice isolation. Try it for free here or integrate it with our [SDK](https://docs.ai-coustics.com/sdk/overview).
+- [Finch](https://ai-coustics.com/2025/06/12/introducing-finch-2-the-new-aicoustics-model-for-studio-quality-speech/): A larger, subtractive model focused on voice isolation and noise removal. Finch removes only unwanted parts of the signal, preserving the original speech. Access it via our [API](https://docs.ai-coustics.com/api-reference/quickstart).
+- [Lark](https://ai-coustics.com/2025/07/29/lark-2-next-generation-reconstructive-speech-enhancement/): Our most advanced, generative model for universal speech enhancement. Lark not only removes noise but also reconstructs and restores lost details in audio, keeping the speaker’s voice intact. Also available through our [API](https://docs.ai-coustics.com/api-reference/quickstart).
 
-You get 120 minutes of free API trial usage. Learn more about our technology and capabilities at [ai-coustics](https://ai-coustics.com/).
+**How to Use**
 
-**How to Use:**
+Upload or record a noisy speech sample to enhance its quality. You can also add background noise and adjust the enhancement level to your preference. Quail is free to use here; for Finch and Lark, obtain an API key from our website (includes a 120-minute free trial). Explore preprocessed examples below to preview the capabilities of each model.
 
-Upload or record a noisy speech sample to enhance its quality. Optionally, add background noise to the input. Adjust the enhancement level to control enhancement you want to apply. You can try Quail for free; to use Lark 2 and Finch, you'll need an API key. Alternatively, preview enhancement results instantly by selecting one of the preprocessed examples below.
-