Image recognition

By Liquid AI

LFM2.5-VL-450M is Liquid AI’s compact vision-language model for structured visual intelligence from edge to cloud. It is built to turn image streams into grounded, actionable outputs in real time, adding object grounding, better instruction following, multilingual image understanding, and function calling support while staying efficient enough for edge hardware.

🔍Image interpretation 🔍Image recognition 📜OCR

NewImage

Released 2mo ago

Gen 1

WildDet3D

By Ai2

WildDet3D is Ai2’s open model for monocular 3D object detection from a single RGB image. It predicts full 3D bounding boxes, including position, size, and orientation in metric coordinates, and supports flexible prompts such as text queries, point clicks, and 2D boxes. It is designed for open-world spatial understanding across varied cameras and can also use optional depth signals when available.

🧩Spatial intelligence 🔍Image recognition 👁️Computer vision assistance

New3d

Released 2mo ago

Gen 3

Falcon Perception

By Technology Innovation Institute

Falcon Perception is TII UAE’s 0.6B early-fusion vision-language model for open-vocabulary grounding and instance segmentation. Given an image and a natural language query, it can return zero, one, or many matching objects with pixel-accurate masks, making it suited for promptable object selection and dense visual localization tasks.

🖼️Image segmentation 🔍Image recognition 🔍Image querying 🔍Object identification

NewMultimodal

Released 2mo ago

Gen 3

Wholembed v3

By mixedbread ai

Wholembed v3 is Mixedbread’s unified omnimodal, multilingual late-interaction retrieval model built for state-of-the-art search across languages and modalities.

🔍Information retrieval 🌐Text translation 🔍Image recognition 🔍Image search

Multimodal

Released 3mo ago

Gen 7

OmniScient Model

By ByteDance

OmniScient Model (OSM) is an open-ended visual recognition approach that predicts free-form class labels for visual entities without requiring a predefined vocabulary at test time.

🔍Image recognition 🖼️Image segmentation 🔍Object identification 👁️Computer vision assistance

Text

Released 3mo ago

Gen 3

Carbon Robotics Large Plant Model

By Carbon Robotics

Large Plant Model (LPM) is Carbon Robotics’ vision model for agriculture, trained on 150M labeled plants to recognize crops and weeds in many climates, powering Carbon AI, LaserWeeder and AutoTractor for precise autonomous weed control.

🌿Plant identification 🔍Image recognition

Multimodal

Released 4mo ago

Gen 4

VIGA

By Fugtemypt123

VIGA is a vision-as-inverse-graphics agent that rebuilds a single image as an editable 3D Blender scene, alternating generator and verifier roles with interleaved multimodal reasoning to capture objects, layout, physics and interactions.

🌍3D images 🔍Image recognition 🎥YouTube thumbnails

Image

Released 4mo ago

Gen 4

SHARP

By Apple

SHARP is Apple’s monocular view-synthesis model that regresses a 3D Gaussian scene from one photo in under a second on a standard GPU, enabling real-time, photorealistic nearby views with metric camera motion.

🖼️Image generation 🔍Image recognition

Image

Released 6mo ago

Gen 4

Precision V2

By Generative Suite

Precision V2 refines V1 with cleaner micro-texture and steadier small-text legibility at similar speed.

🔍Image upscaling 🖌️Image editing 🔍Image recognition

Image

Released 7mo ago

Gen 4 Hunyuan

HunyuanWorld Mirror

By Tencent

HunyuanWorld Mirror is a scene-reconstruction and world-modeling system. It turns photos and videos into a consistent digital twin that you can explore, edit, and render, with export to common 3D formats for simulation, virtual production, and design.

🖼️Image generation 🔍Image recognition

Image

Released 7mo ago

Gen 4

Crystal Upscaler

By ClarityAI

Crystal Upscaler is an image super-resolution and enhancement model that enlarges 2x to 8x while restoring detail, reducing noise, and fixing compression artifacts. It works on photos, renders, anime, and UI art with controllable sharpness and texture preservation.

🔍Image upscaling 🖌️Image editing 📝Social media bios 🔍Image recognition

Image

Released 7mo ago

Gen 3

FlowRVS

By xmz111

FlowRVS is a referring video object segmentation method that learns a text-conditioned continuous flow to deform a video’s spatiotemporal representation into the target object mask.

🖼️Image segmentation 🔍Image recognition 🎥Video analysis 👁️Computer vision assistance

Multimodal

Released 8mo ago

Gen 3

CameraTrapAI

By Google

Google-provided AI models for classifying wildlife species in camera trap images.

🔍Image recognition 🧝‍♀️Fantasy images

Multimodal

Released 8mo ago

Gen 4

Pixel Perfect Depth

By Xiaomi

Pixel-Perfect Depth is a monocular depth estimation model that uses pixel-space diffusion transformers to predict high-quality, flying-pixel-free depth maps for dense point clouds, accepted at NeurIPS 2025.

🖼️Image generation 🔍Image upscaling 🔍Image recognition

Image

Released 8mo ago

Gen 4 Hunyuan

HunyuanImage 3.0

By Tencent

HunyuanImage 3.0 is Tencent’s next-gen text-to-image model. It delivers sharper detail, stronger style and identity consistency, improved typography, and precise, in-place editing—built for fast iteration from concept to production-ready visuals.

🖼️Image generation 🖌️Image editing 💡Coding help 🔍Image recognition

Image

Released 8mo ago

Gen 3 Qianfan

Qianfan-VL-3B

By Baidu

Qianfan-VL-3B is Baidu’s lightweight VLM for cost-sensitive, real-time multimodal apps. It processes images plus text and returns grounded answers with basic OCR and layout understanding, long context, tool/function calling, and JSON outputs—optimized for speed and efficiency.

🏭Manufacturing 🖼️Image to text 🔍Image recognition

Text

Released 8mo ago

Gen 3 Qianfan

Qianfan VL 70B

By Baidu

Qianfan-VL 70B is Baidu’s large vision-language model on the Qianfan platform. It ingests images (docs, charts, screenshots, photos) with text and produces grounded answers, featuring strong OCR and layout understanding, long context, tool/function calling, streaming, and reliable JSON outputs for multimodal RAG and enterprise apps.

📜OCR 🖼️3D image generation 🎬Video dubbing 🔍Image recognition

Text

Released 8mo ago

Gen 3

Moondream 3 Preview

By Moondream

Moondream 3 Preview is a compact frontier-oriented vision-language model built for fast visual reasoning, grounding, OCR, object detection, pointing, and structured output. It uses a 9B MoE architecture with 2B active parameters and extends context length to 32K, aiming to deliver strong real-world vision performance while staying efficient and inexpensive to run.

🔍Image interpretation 🔍Image recognition 🖼️Image segmentation 📜OCR

Multimodal

Released 8mo ago

Gen 3 Command

Command A Vision

By Caldera Labs

Command A Vision is Cohere’s multimodal instruction model that pairs text and image understanding. It accepts images plus text prompts and outputs structured, step-by-step text answers. It’s tuned for enterprise workflows like document OCR, chart/diagram reasoning, screenshot/UI analysis, and tool or function calling.

📜OCR 🖼️Image to text 🔍Image recognition

Text

Released 10mo ago

Gen 3 Kanana

Kanana 1.5

By Naver

Kanana-1.5-v-3B is a 3B-parameter vision–language model in Kakao’s Kanana line. It can process both images and text prompts, outputting grounded answers in natural language or structured JSON. It’s optimized for lightweight multimodal assistants and enterprise applications that need efficiency with visual reasoning.

📜OCR 🔍Image recognition

Text

Released 10mo ago

Gen 4 Earth

Earth 2 FourCastNet 3

By NVIDIA

Earth-2 FourCastNet 3 is a geometric ML global ensemble model that respects spherical Earth geometry to deliver fast, probabilistic medium- to subseasonal forecasts, outperforming leading numerical ensembles at much lower cost.

❤️Empathetic conversations 🔍Image recognition 💥Bdsm education

Image

Released 10mo ago

Gen 3

moondream 2

By Moondream

Moondream 2 is a small open vision-language model designed to run efficiently across many environments. It is the previous-generation Moondream model, released under Apache 2.0, and is positioned as a lightweight image-text model for practical multimodal use where smaller size and deployability matter more than maximum frontier scale.

🖼️Image generation 🔍Code reviews 🔍Image recognition

Multimodal

Released 11mo ago

Gen 4

Hunyuan HY-T1

By Alibaba

Hunyuan T1 is Tencent’s deep reasoning model positioned for stronger structured reasoning and long-context analysis.

🖼️Image generation 🔍Image recognition

Image

Released 1y ago

Gen 3 Qwen

Qwen 2.5-VL-72B

By Alibaba

Qwen 2.5-VL-72B is Alibaba’s flagship open-weight vision-language model. It takes images (docs, charts, screenshots, photos) plus text and answers in text, with strong OCR, layout understanding, and multi-image reasoning. It supports long context, function/tool calling, and reliable JSON outputs—ideal for multimodal RAG, agents, and enterprise workflows.

🏭Manufacturing 🖼️Image to text 🔍Image recognition

Text

Released 1y ago

Gen 4

Moondream 0.5B

By Moondream

Moondream 0.5B is a tiny open-source vision-language model built for edge devices and mobile platforms. With only 0.5B parameters, it is positioned as the world’s smallest VLM, designed for fast lightweight deployment on constrained hardware while still supporting practical real-world visual tasks.

🔍Image interpretation 🔍Image recognition

Image

Released 1y ago

Gen 3 Gemma

PaliGemma 2

By Google DeepMind

PaliGemma 2 is Google’s upgraded open vision-language model family based on Gemma 2, available in 3B, 10B, and 28B sizes.

🔍Image recognition

Text

Released 1y ago

Gen 4

Zen

By Freepik Company

Zen aims for minimalist, calm compositions with natural lighting and restrained color.

🖼️Image generation 🔍Image recognition 🔄Website conversion 📐Blueprints

Image

Released 1y ago

Gen 3

NV-CLIP

By NVIDIA

NV-CLIP is NVIDIA’s CLIP-style vision–language encoder that maps images and text into a shared embedding space for visual search, cross-modal retrieval, and zero-shot classification. It’s optimized for NVIDIA GPUs and easy to deploy at scale.

🎨NFT art 🔊Text to speech 🔍Image recognition

Text

Released 1y ago

Gen 6 LLaMA

Llama 3.2

By Meta Platforms

Includes lightweight text models (1B, 3B for edge/mobile, 128k context) and vision models (11B, 9...

📷Images 🔍Image recognition

Text

Released 1y ago

Gen 4

OmniParser

By Microsoft

OmniParser is a comprehensive method for parsing user interface screenshots into structured and easy-to-understand elements, which significantly enhances the ability of GPT-4V to generate actions that can be accurately grounded in the corresponding regions of the interface.

🔍Image recognition 🖼️Image generation

Image

Released 1y ago

Gen 3 Palmyra

Palmyra Vision

By Writer Engineering

Palmyra Vision is Writer’s multimodal LLM that takes images as input and generates text output. It can extract text from images (including handwriting), interpret charts/graphs/diagrams, classify objects, and answer questions about visual content—all aimed at enterprise workflows.

🖼️Image to text 🔍Image recognition 🖼️Image descriptions

Text

Released 2y ago

Gen 4

Lightning XL

By Leonardo Interactive

Lightning XL is a speed optimized SDXL checkpoint that produces strong images in very few steps, ideal for rapid iteration.

🖼️Image generation 🔍Image upscaling 🔍Image recognition

Image

Released 2y ago

Gen 4

Ultralytics YOLO

By Ultralytics

Ultralytics YOLO is a family of real-time computer-vision models for detection, segmentation, classification, pose, and tracking, designed to be fast, accurate, and easy to deploy across edge and cloud.

🔍Image recognition

Image

Released 3y ago