AI Foundation Model Registry

Registered foundation runtimes against which prompts are evaluated and benchmarked (10).

GPT-4oOpenAI

Omni flagship model with high-speed reasoning, coding, and vision support.

128k tokens
Claude 3.5 SonnetAnthropic

State-of-the-art reasoning, deep coding execution, and nuanced editorial prose.

200k tokens
Claude 3 OpusAnthropic

Deep conceptual synthesis, academic literature assimilation, and complex thought.

200k tokens
Gemini 1.5 ProGoogle

Massive multi-million token context window ideal for codebase auditing and long books.

2M tokens
DeepSeek R1DeepSeek

Open-weights reasoning model with explicit verification and chain-of-thought traces.

64k tokens
Llama 3.3 70BMeta

Leading open-weights model balancing reasoning speed and open enterprise deployment.

128k tokens
Mistral LargeMistral AI

Top-tier European multilingual foundation model specializing in reasoning and code.

128k tokens
Qwen 2.5 72BAlibaba

High-performance mathematical and multilingual reasoning model with code mastery.

128k tokens
Midjourney v6Midjourney

Photorealistic, cinematic image rendering and typographic prompt comprehension.

Visual Gen
Stable Diffusion 3.5Stability AI

Multimodal diffusion transformer for customizable local and cloud visual generation.

Visual Gen