Model Square

Flagship hybrid reasoning model for coding and complex tasks

Pinned Claude 3.7 Sonnet snapshot for coding and complex tasks

Pinned Claude 3.5 Sonnet revision 2 snapshot

Claude 3.5 Sonnet revision 2 for coding and general reasoning

Fast reasoning model optimized for latency and cost

Pinned GPT-4o multimodal snapshot

Flagship multimodal model; provider alias tracks the current snapshot

Fast multimodal model with a 1M-token context window

Advanced reasoning with a 1M-token context window

Efficient DeepSeek V3 general-purpose model

Open reasoning model; response time varies with reasoning depth

Qwen flagship model for complex multilingual tasks

Open coding model for responsive code generation

Fast everyday assistant

High-capability reasoning

Fast and economical

Affordable multimodal model

Fast text-to-image generation

High-quality text-to-image generation

Advanced image generation

Text-to-image generation

OpenAI image generation model

Creative image generation model

Advanced reasoning

Efficient reasoning

Advanced reasoning model

Fast reasoning model

Large context coding model

Efficient coding model

Low-cost large context model

Reliable high-capability chat model

Most affordable Gemini

Fast multimodal assistant

Lightweight Gemini model

Strong coding and chat

Strong open-source coding model

Reasoning-focused model

Specialized code generation model

Open model for general tasks

Large open-source model

Strong open-source model

Long context open-source

European flagship model

Fast and affordable

Specialized code model

Multilingual open model

Reasoning and multilingual model

Strong bilingual model

Strong bilingual model

Tool-use and retrieval model

Search-aware assistant

Conversational model with broad knowledge