Try Hy3 online Demo. Hy3 is Tencent’s open 295B MoE model with 21B active parameters, 256K context, Apache 2.0 license, and agent-focused features.

• #Agent
• #ChatBot

Ornith is an open-source AI coding model family from DeepReinforce. Learn what it does, model sizes, license, features, and use cases.

• #Coding

Boogu is an Apache-2.0 image generation and editing model family with text-to-image, fast Turbo, fp8, and editing variants.

• #Image Edit

VoxCPM 2 is an open-source TTS model for multilingual speech, voice design, cloning, and 48kHz audio. See features, setup, and limits.

• #Text to Speech

Ideogram 4 is an open-weight text-to-image model with JSON prompts, strong text rendering, 2K output, and API access for teams.

• #Design
• #Text to Image

MiniMax M3 is an open-weight multimodal AI model with 1M context, sparse attention, coding strengths, API access, and local deployment.

• #ChatBot

Pixal3D turns reference images into 3D assets. Learn what it is, key features, pricing, setup options, and practical use cases.

• #Image to 3D

North Mini Code is Cohere’s open-weights coding model for agentic software engineering, terminal tasks, and local code generation.

• #Agent
• #Coding

DiffusionGemma is Google DeepMind’s open text-diffusion Gemma 4 model. Learn its features, use cases, limits, and where to try it.

• #Recommand

Bernini-R is ByteDance’s open-source video renderer for Bernini, supporting image, video, and reference-guided generation and editing.