Description:
ModLens is a vision plugin for DeepSeek Harness that lets text-only AI models read pasted images, returning structured JSON evidence (OCR, layout, semantics) and enabling vision-enabled outputs across multiple hosts.
Keep Calm and Read the Friendly Manual :-)
Description:
ModLens is a vision plugin for DeepSeek Harness that lets text-only AI models read pasted images, returning structured JSON evidence (OCR, layout, semantics) and enabling vision-enabled outputs across multiple hosts.
Description:
Cua is an open-source platform for agent-ready sandboxes across macOS, Windows, Linux, and containers. It provides a unified driver, sandbox API, and benchmarks for building and evaluating autonomous agents.
License: Dual license: Open Source License (AGPL-3.0) and Enterprise License (Commercial)
Description:
Kodus AI is a monorepo for AI-powered code review, enabling users to run reviews with BYOK model providers, choose from multiple LLMs, apply Kody Rules, and monitor outcomes with Cockpit. Self-host or use Kodus Cloud.
Description:
OpenDroid is an autonomous AI agent for Android that plans, executes, and verifies tasks on-device, enabling hands-free automation with multi-LLM support and on-device model management.
Description:
FaceSwap is a Python tool that uses deep learning to recognize and swap faces in photos and videos, featuring multiple models and GUI components.
Description:
Whisper is a general-purpose speech recognition model by OpenAI, capable of multilingual transcription, translation, and language identification. Transformer-based, trained on diverse audio data.
Description:
Open-source diffusion-transformer TTS framework delivering fluent, faithful speech with Flow Matching. Includes F5-TTS Base, E2-TTS, multi-speaker support, CLI and Gradio interfaces, and Docker-ready deployment.
Description:
A deep learning toolkit for Text-to-Speech with pretrained models, training utilities, and data tools for building speech synthesis across many languages.
Description:
End-to-end web-based interface for retrieval-based voice conversion using Open Source RVC models and Torch-based ML pipelines.
Description:
Cognee is an open-source AI memory platform that ingests data into a self-hosted knowledge graph, enabling persistent memory for agents across sessions with vector search and ontology grounding.