feat: add highly visible Call to Action prompting users to like the HF space

fix(core): migrate extraction to PCM-WAV to bypass libmp3lame missing codec and enforce 25s limit on cloning endpoints

fix: critically restore ssr_mode=False to bypass localhost environment health check hanging in Gradio

fix: add dynamic ffmpeg auto-cropping to bypass Mistral API 30s ref_audio limitation

docs: clarify mistral billing requirement

refactor: decouple API functions to core.py

feat: split out application backend into reusable core.py module for offline local developers

chore: remove bloated leaderboard src folder

chore: delete unneeded template file .pre-commit-config.yaml

chore: delete unneeded template file pyproject.toml

chore: delete unneeded template file Makefile

fix: completely remove backdrop-filter which causes webkit positioning bugs and pushes fixed elements

fix: completely obliterate unsafe global input CSS overrides breaking Gradio dropdown positioning logic

fix: disable padding override that misaligns dropdown and style drop items

fix: correct webkit text-fill inheritance causing blank badge

fix: unclip dropdown menus by setting container overflow to visible

fix: remove demo.load to prevent Gradio 6 tab component race condition freezing UI

fix: import tempfile and stabilize UI CSS rendering in Gradio 6

fix: populate voices via lazy load, add direct mp3 audio curl bypass for YT blocks, add clone badge

fix: correct gradio 6 launch parameters to prevent localhost crash

style: completely overhaul UI aesthetic (glassmorphism) and fix yt-dlp youtube blocks

fix: explicitly set server_name=0.0.0.0 and server_port=7860 to avoid Gradio 6 healthcheck ValueError on HF spaces

feat: add ability to clone voice natively from URLs (YouTube, TikTok) using yt-dlp

build: add yt-dlp

build: add ffmpeg to debian packages

chore: add HF discovery tags to metadata so the space appears in search

fix: prevent user-created voices from being leaked globally while retaining official presets

feat: upgrade Voice ID textbox into a dynamic Dropdown that natively fetches mistral voices across the UI

feat: add List My Voices button to execute client.audio.voices.list() dynamically

fix: add UI error when no voice is provided instead of throwing 404 error using a hardcoded fallback id

fix: revert voice back to voice_id for python sdk kwargs

fix: resolve TTS base64 error and fix voice SDK parameter

fix: correct mistral import path for mistralai>=2.3.2

feat: add zero-shot ref_audio directly to TTS tab

fix: upgrade mistralai to >=2.3.2 to get audio API

fix: gradio 6 breaking changes in app.py

fix: upgrade to gradio 6.12.0 (latest, no HfFolder bug)

fix: only mistralai==1.2.0, let gradio 5 handle huggingface_hub

fix: pin huggingface_hub==0.25.0 and mistralai==1.2.0 explicitly

fix: pin huggingface_hub<0.26.0 to restore HfFolder for Gradio 4.x

fix: remove gradio from requirements, let sdk_version 5.0.1 handle it

fix: remove gradio from requirements, let sdk_version 5.0.1 handle it

fix: upgrade to gradio 5 to fix HfFolder import error

fix: upgrade to gradio 5 to fix HfFolder import error

fix: pin python_version to 3.11 to fix audioop crash on Python 3.13

Upload README.md

Upload requirements.txt

Upload app.py

initial commit

Duplicate from gradio-templates/leaderboard