Neural Cosmos

AI Models

A gallery of frontier models — language, vision, and multimodal systems driving the AI landscape.

Featured

GPT-4

Large multimodal model that exhibits human-level performance on various benchmarks.

OpenAIReleased 2023language · lg

Gemini 1.5

multimodal

Google's most capable model with a 1M token context window.

Google DeepMind2024

Claude 3.5

multimodal

State-of-the-art reasoning, vision, and coding with near-human accuracy.

Anthropic2024

Llama 3

language

Open mixture-of-experts model optimized for speed and efficiency.

Meta AI2024

GPT-4o

multimodal

Omni model optimized for fast, multimodal interaction.

OpenAI2024

Mistral Large

language

High-performance open model for reasoning-heavy workloads.

Mistral AI2024

Titan Image

vision

Diffusion-based image generation with fine-grained control.

Amazon2023

Claude Vision

vision

Advanced visual understanding with grounding and OCR.

Anthropic2024