Open source

Pika releases open-source H3 Flex for joint audio-video generation

Pika released H3 Flex, a derivative model of MiniMax H3, with open weights and inference code. The model jointly generates video and audio via six-step denoising, supporting multimodal inputs and high-resolution enhancement for developers.

Summarized by AI from one outlet

Source

Latest

  1. Open source

    Meituan open-sources LongCat-2.5 model

    Meituan LongCat open-sourced its native multimodal model LongCat-2.5 under the MIT license. The model has 1.6 trillion parameters, supports 1M context, and is optimized for domestic compute chips to enhance agentic and visual coding tasks.

    IT之家

  2. Open source

    Reason Machines Open-Sources Aspen, a Git Server for AI Agents

    Reason Machines has open-sourced Aspen, a self-hosted Git server designed for AI agents. It supports the OpenAI Pando design and achieved 567 pushes per second on a single 8-core server, enhancing AI development infrastructure.

    MarkTechPost

  3. Open source

    0sec Labs Open-Sources AI Security Harness 0

    0sec Labs has open-sourced 0, an AI security harness for code review, pentesting, and reverse engineering. It supports execution via browser, CLI, and MCP server, allowing users to customize model providers to enhance local security defenses.

    MarkTechPost

All news → · All rights belong to the original creators