Opus 4.6, GPT 5.3 Codex, StepFun, Qwen3 Coder, new deepfake AIs, new video tools: AI NEWS
Recent AI advancements include models for video animation, 3D rigging, open-source omnimodal interaction, video editing, and superior OCR capabilities, with notable releases from Enthropic, OpenAI, ZAI, and Tencent.
MAIN POINTS FROM TRANSCRIPT
- New AI models can transfer movement and control camera in animations.
- Open-source models now offer omnimodal interaction with text, voice, and video.
- ZAI's GLM OCR outperforms other models in text and data parsing from images.
- Tencent's Interact Avatar enables text-prompted object interaction in animations.
TAKEAWAYS
- Enthropic and OpenAI released advanced proprietary AI models, but open-source options are competitive.
- GLM OCR excels in speed and accuracy, handling multiple languages and formats efficiently.
- Open-source coding agents rival top proprietary models in performance.
- AI advancements include real-time video editing and interaction capabilities.