Safeguard against Gradio ffmpeg postprocessing conflicts by returning a unique copy of the output video and removing format='mp4' constraint

Fix AssertionError in Gradio when reusing inputs by replacing shutil.move with shutil.copy

Fix numpy inhomogeneous array error by explicitly casting to float

Fix deprecated numpy types (np.float -> float, np.int -> int, etc) that break on newer numpy versions

Fix Python 3 TypeError by raising ValueError and lower face detection threshold to 0.80

Make facexlib patch robust to save_dir argument

Fix permissions errors by patching weight download paths

Return error log

Enable verbose error reporting

Add traceback to error

Pin pydantic version to fix Gradio 5 bool argument crash

Patch torchvision compatibility for basicsr and remove unused files

Upgrade Gradio to v5.9.1 to resolve huggingface_hub HfFolder import error

Restore missing Hugging Face Space configuration in README.md

Completely rewrite the Space to a Talking Avatar Converter powered by SadTalker and ZeroGPU

Fix Wan-AI model repository name to use Diffusers-compatible version

Use DiffusionPipeline instead of AutoPipelineForText2Video

Fix ZeroGPU ImportError by adding missing Wan-AI dependencies and surface precise errors

Complete SEO overhaul and brand rename to Ultimate AI & OSINT Toolkit

Switch to Wan-AI/Wan2.1-T2V-1.3B local ZeroGPU text-to-video generation

Switch Video Batch Generator to use Hugging Face Inference API instead of local ZeroGPU

Fix LTX-Video ValueError: change height from 720 to 704 to be a multiple of 32

Upgrade Text-to-Video engine to Lightricks/LTX-Video for native 720p 24fps generation

Fix ZeroGPU illegal duration: lower requested duration to 120s

Fix requirements.txt encoding corruption

Add local AI Video Batch Generator using zeroscope_v2_576w on ZeroGPU

Implement global AI model provider to enhance Deepfake and EXIF tabs with intelligent summaries

Refactor UI: Make AI Model Provider dropdown global across all tabs

Update Cohere model identifier to command-r-plus-08-2024

Update fallback provider model identifiers to avoid 404 and 410 errors

Fix environment variable names to match .env file (FALLBACK_*)

Restore call_ai_api implementation

Remove Hugging Face default API and make Groq the primary default

Fix function corruption: restore compare_real_vs_suspect and compare_two_gpu

Fix requirements.txt encoding to remove null bytes

Integrate multiple AI API fallback providers from .env into OSINT Robolawyer

Update .env

Update .env

Update .env

Update .env

Update .env

Update .env

Update .env

Update .env

Create .env

Update monkeypatch to target _json_schema_to_python_type directly to avoid APIInfoParseError

Add gradio_client get_type monkeypatch to fix boolean schema TypeError on startup

Remove strict gradio>=5 constraint to fix schema parsing issues

Fix Gradio ChatInterface type for Gradio 5 compatibility

Fix Gradio ChatInterface schema bug