TAAFT
Free mode
100% free
Freemium
Free Trial
Deals
Companies
ByteDance
Models
Depth Anything 3

Depth Anything 3

By ByteDance
Depth Anything 3 uses a single plain transformer backbone and a depth-ray representation to cover monocular depth, multi-view depth, pose-conditioned depth, pose estimation, and direct 3D Gaussian estimation for novel view synthesis, with released models and tooling.
New Image Gen 4
Released: November 13, 2025

Overview

Depth Anything 3 is a spatial reconstruction model that predicts consistent geometry from arbitrary visual inputs, with or without known camera poses.

🧩Spatial intelligence🌐2D to 3D image conversion👁️Computer vision assistance

About ByteDance

ByteDance is a multinational technology company known for its content platforms, including TikTok and Douyin.

Industry: Internet
Company Size: 10001+
Location: Beijing, CN
Website: bytedance.com
View Company Profile

Tools using Depth Anything 3

No tools found for this model yet.

Related Models

  • Gen 7
    Doubao-1.5-Pro
    ByteDance's flagship LLM upgrade aimed at advanced reasoning tasks. Offers extremely long context...
    🎨NFT art📝Writing📷Images
    Text
    Released 1y ago
  • Gen 4
    Earth 2 CorrDiff
    Earth-2 CorrDiff is a generative correction-diffusion model that downscales coarse forecasts and climate projections into kilometer-scale, bias-corrected fields up to 500x faster and around 10,000x more energy efficient than traditional methods.
    🖼️Image generation🔍Image upscaling
    Image
    Released 1y ago
  • Gen 4
    video SALMONN 2
    video-SALMONN 2 is an audio-visual large language model from Tsinghua and ByteDance that uses video frames and sound to generate rich captions and answers, reaching state-of-the-art on many audio-visual QA and video understanding benchmarks.
    🎥Videos❓Answers
    Text
    Released 9mo ago
Last updated: March 12, 2026
0 AIs selected
Clear selection
#
Name
Task