Anthropic's most capable model with best-in-class safety and reasoning.

Chinese open-source model rivaling GPT-4 at a fraction of the cost.

OpenAI's most advanced language model with enhanced reasoning, multimodal understanding, and agentic capabilities.

Fast and cost-effective model with strong reasoning capabilities and 1M token context window.

Google's most powerful multimodal AI model with native search integration.

Meta's open-source LLM with competitive performance and full customizability.

Cost-effective multimodal model for fast processing of image, video, and text inputs.

Highly capable multimodal model offering best accuracy-cost-speed balance for enterprise workloads.

On-device and server AI models powering Apple Intelligence features across the Apple ecosystem.

Best multilingual open model supporting 23 languages with strong cross-lingual transfer.

Chinese enterprise model with strong business application focus.

Specialized conversational AI platform enabling users to create and chat with customizable AI characters.

Fastest model in the Claude 3.5 family, optimized for speed and cost efficiency while maintaining strong performance.

Anthropic's balanced model offering strong performance at moderate cost.

Anthropic's largest and most capable model with exceptional reasoning, analysis, and creative writing abilities.

Advanced reasoning model with strong coding and analysis capabilities, balancing performance and speed.

Dedicated code generation model from Mistral optimized for programming tasks across 80+ languages.

Enterprise-focused model with strong RAG, tool use, and agentic capabilities for business applications.

Enterprise-focused model optimized for RAG and business applications.

Cohere's updated flagship model with improved RAG capabilities and enterprise-grade reliability.

Cohere enterprise-focused LLM optimized for business applications with strong retrieval-augmented generation, multilingual support, and grounded responses.

Compact enterprise RAG-optimized model designed for retrieval-augmented generation workflows.

Open-source mixture-of-experts model setting new benchmarks for open models in efficiency and quality.

Second-generation reasoning model from DeepSeek with improved mathematical and coding capabilities, building on the open-source R1 architecture.

Baidu's flagship model with deep Chinese internet knowledge.

UAE's Technology Innovation Institute open-source model with strong benchmark results.

Chinese bilingual model with strong reasoning and tool-use capabilities.

OpenAI's optimized multimodal model balancing speed, cost, and capability.

Compact and affordable multimodal model offering strong performance at a fraction of GPT-4o cost.

OpenAI's latest GPT iteration with improved reasoning, reduced hallucinations, and expanded multimodal capabilities.

Google's lightweight open model family for efficient deployment.

Latest compact open-source model family from Google, optimized for efficiency and on-device deployment.

Enterprise-grade open-source LLM built for business applications with strong governance features.

xAI's witty, unfiltered model with real-time X (Twitter) data access.

xAI's updated large language model with real-time knowledge, humor, and deep reasoning capabilities.

Conversational AI model focused on empathetic, personal interactions with strong emotional intelligence.

Chinese research model with strong tool use and agent capabilities.

Hybrid SSM-Transformer architecture for efficient long-context processing.

Long-context specialist with industry-leading 2M token context window.

Moonshot AI's long-context model with enhanced reasoning and 2M token context window.

Meta's powerful Llama 4 variant designed for complex reasoning tasks with mixture-of-experts architecture.

Meta's efficient Llama 4 variant optimized for speed and low-resource deployment while maintaining strong performance.

Chinese AI model with strong multilingual capabilities and competitive performance on global benchmarks.

Mistral code-specialized model using the Mamba state-space architecture for efficient long-context code generation with linear scaling.

Mistral's latest mid-tier model balancing performance and efficiency for enterprise applications.

Efficient Mixture of Experts model with strong performance per compute.

NVIDIA's model optimized for GPU inference with enterprise-grade performance.

Fully open model with transparent training data and methodology.

Search-augmented language model providing real-time, cited answers by combining LLM reasoning with live web search.

Action-oriented model designed for computer use and automation.

Microsoft's compact powerhouse proving small models can punch above their weight.

Alibaba's multilingual model excelling in Chinese and English tasks.

Vision-language model with strong visual understanding.

Multimodal model with native video understanding capabilities.

Open-source enterprise model optimized for SQL generation, coding, and data analytics tasks.

Specialized open-source coding model trained on permissive code.

Microsoft's enterprise NLG model for business content generation.

Instruction-following specialist with Evol-Instruct training methodology.

Enterprise AI model excelling at business writing, compliance, and domain-specific content generation.

Fast inference model balancing speed and quality for production use.

OpenAI advanced reasoning model with breakthrough performance on math, science, and coding benchmarks, using extended chain-of-thought for complex problem solving.

OpenAI cost-efficient reasoning model balancing strong analytical capabilities with faster inference and lower costs compared to full reasoning models.