jianzhnie/awesome-text-to-video

What it solves

Keeping up with the fast-moving text-to-video (T2V) field is hard: new commercial tools, open-source models, and research papers appear constantly, and older ones get discontinued. This project solves that by gathering everything into one visually organized, up-to-date list, so you can quickly compare options, discover what's available, and find the right tool for your use case.

How it works

The list is structured into clear sections. It covers commercial products (generative models, AI avatar tools, all-in-one creative platforms, and discontinued ones), featured product cards with quick summaries, related tools, open-source models and toolkits with details like size and minimal GPU memory, research papers sorted by year, and datasets & benchmarks. A "How to Choose" table gives direct recommendations based on common goals (e.g., professional filmmaking, budget motion realism, corporate training), and a "What's New" section highlights major shifts and trends updated as of 2026.

Who it's for

Anyone involved in video creation or AI research: content creators, filmmakers, marketers, corporate trainers, developers, and researchers. It helps both beginners looking for an easy entry point (like a budget or avatar tool) and experts who need to track open-source models, papers, and benchmarks.

Highlights

• Covers both commercial and open-source landscape, including discontinued products (Sora, Haiper) to avoid dead ends.
• Direct comparison tables for products, models, and tools, with metrics like max output resolution, size, VRAM, and best-use cases.
• Research papers organized by year (2022–2026 and before), making it easy to follow academic progress.
• Includes a curated "How to Choose" guide and featured product cards for at-a-glance decisions.
• Mentions practical open-source models (Wan 2.2, HunyuanVideo, LTX-Video, etc.) with code links and system requirements.