AI Models

Claude Fable 5.2 is a model from Anthropic designed for demanding reasoning, advanced coding, vision, and long-running agentic work. It shares its underlying model with Claude Mythos 5.2 and includes safeguards for general use.

Picks automatically from: GPT-6 Luna · GPT-6.1 Terra · GPT-6.1 Sol · GPT 6 Astra

Picks the right OpenAI model for every message: GPT-6 Luna for everyday requests, GPT-6.1 Terra for visual and computer-use tasks, GPT-6.1 Sol for demanding work, and GPT-6 Astra only for the most complex problems. You get the quality the task needs without paying for a large model where a smaller one is enough, and you never have to switch models yourself. Each answer shows which model wrote it.

A powerful model optimized for precise instruction following, coding, and tool use. With a context window of up to 1 million tokens, it is well suited for processing extensive documents, complex tasks, and large codebases.

A smaller and faster variant of GPT‑4.1 offering a balanced combination of performance, speed, and operating cost. It is ideal for everyday text processing, coding, automation, and handling high volumes of requests.

The fastest and most cost-efficient model in the GPT‑4.1 family, designed for straightforward, high-volume tasks. It is ideal for classification, text completion, information extraction, and other applications requiring minimal latency.

A lightweight and cost-efficient model in the GPT‑5.6 family, optimized for processing high volumes of requests quickly. It is well suited for smaller everyday tasks, automation, text processing, and applications where low cost and fast response times are essential.

The most advanced model in the GPT‑5.6 family, designed for demanding knowledge work, complex reasoning, and coding. With a context window of over 1 million tokens, it can handle large projects, extensive documents, and sophisticated agentic workflows.

The most advanced model in the GPT‑5.6 family, designed for demanding knowledge work, complex reasoning, and coding. With a context window of over 1 million tokens, it can handle large projects, extensive documents, and sophisticated agentic workflows.

A versatile model in the GPT‑5.6 family that provides a balanced combination of intelligence, speed, and operating cost. It is ideal for everyday professional work, analysis, coding, and automation at scale.

A versatile model in the GPT‑5.6 family that provides a balanced combination of intelligence, speed, and operating cost. It is ideal for everyday professional work, analysis, coding, and automation at scale.

GPT-6 Astra is OpenAI's most capable model, built for the hardest end-to-end work. Use it for complex reasoning, coding, computer use, research, and document creation.

GPT-6 Astra is OpenAI's most capable model, built for the hardest end-to-end work. Use it for complex reasoning, coding, computer use, research, and document creation.

Flagship image model with the highest quality and detail. Use for final assets and complex compositions.

A lightweight and cost-efficient model in the GPT‑6 family, optimized for processing high volumes of requests quickly. It is well suited for smaller everyday tasks, automation, text processing, and applications where low cost and fast response times are essential.

A fast and affordable model in the GPT‑6 family that brings much of GPT‑6 Astra's strength in coding, factual accuracy, computer use and professional work to everyday use. A strong choice for complex tasks, coding and agentic workflows at a lower price.

A fast and affordable model in the GPT‑6 family that brings much of GPT‑6 Astra's strength in coding, factual accuracy, computer use and professional work to everyday use. A strong choice for complex tasks, coding and agentic workflows at a lower price.

A fast and affordable model in the GPT‑6 family that brings much of GPT‑6 Astra's strength in coding, factual accuracy, computer use and professional work to everyday use. A strong choice for complex tasks, coding and agentic workflows at a lower price.

A fast and affordable model in the GPT‑6 family that brings much of GPT‑6 Astra's strength in coding, factual accuracy, computer use and professional work to everyday use. A strong choice for complex tasks, coding and agentic workflows at a lower price.

A fast, efficient model for visual reasoning, precise computer use, and clear, natural writing. A strong choice for matching product photos to catalog entries, interpreting technical drawings, acting on application screens, or drafting and revising text.

Fast, high-quality everyday image generation (recommended as default image generation model)

Most capable GPT Image generation model

Picks automatically from: Claude Haiku 5.5 · Claude Sonnet 5.5 · Claude Opus 5.5 · Claude Fable 5.1

Picks the right Claude model for every message: Claude Haiku 5.5 for quick simple requests, Claude Sonnet 5.5 for most work, Claude Opus 5.5 for hard problems and Claude Fable 5.1 for exceptional ones. You get the quality the task needs without paying for a large model where a smaller one is enough, and you never have to switch models yourself. Each answer shows which model wrote it.

Fable 5 is a Mythos-class model that Anthropic made safe for general use — its capabilities exceed any model they've ever made generally available, and it's state-of-the-art on nearly all tested benchmarks.

Fable 5.1 is a Mythos-class model that Anthropic made safe for general use — its capabilities exceed any model they've ever made generally available, and it's state-of-the-art on nearly all tested benchmarks.

A fast and cost-efficient model optimized for straightforward tasks and high-volume request processing. It is well suited for concise responses, classification, information extraction, and automated conversations.

Fast and cost-efficient. Anthropic's fastest and most capable small model, built for high-volume and speed-sensitive tasks such as summarization, classification, subagents, live customer support and browser use.

A high-performance model designed for complex analysis, advanced reasoning, and high-quality content creation. It excels at expert-level tasks, large-context processing, and multi-step assignments.

A high-performance model designed for complex analysis, advanced reasoning, and high-quality content creation. It excels at expert-level tasks, large-context processing, and multi-step assignments.

A high-performance model designed for complex analysis, advanced reasoning, and high-quality content creation. It excels at expert-level tasks, large-context processing, and multi-step assignments.

An advanced model for highly demanding tasks, offering precise reasoning and reliable handling of complex instructions. It is well suited for strategic analysis, software development, and expert content creation.

An advanced model for highly demanding tasks, offering precise reasoning and reliable handling of complex instructions. It is well suited for strategic analysis, software development, and expert content creation.

The most advanced model in the Opus family, built for tasks requiring maximum quality, accuracy, and depth of reasoning. It is an excellent choice for complex decision-making, expert research, and large-scale projects.

The most advanced model in the Opus family, built for tasks requiring maximum quality, accuracy, and depth of reasoning. It is an excellent choice for complex decision-making, expert research, and large-scale projects.

A practical everyday model from Anthropic that delivers near-frontier intelligence for software engineering, knowledge work and research. It verifies its own work and iterates carefully, which makes it reliable for longer agentic tasks.

Anthropic's flagship model for complex knowledge work, agentic coding and long-running autonomous tasks, performing at the level of Claude Fable 5.1 on most work. It is faster and cheaper to run than Opus 5 and writes clear, well-structured answers that put the most important information first.

Anthropic's flagship model for complex knowledge work, agentic coding and long-running autonomous tasks, performing at the level of Claude Fable 5.1 on most work. It is faster and cheaper to run than Opus 5 and writes clear, well-structured answers that put the most important information first.

A versatile model offering an excellent balance of quality, speed, and operating cost. It is ideal for everyday text processing, data analysis, coding, and workflow automation.

A versatile model offering an excellent balance of quality, speed, and operating cost. It is ideal for everyday text processing, data analysis, coding, and workflow automation.

A powerful and efficient next-generation model for a wide range of professional and creative tasks. It delivers high-quality results in complex reasoning, coding, and large-context processing.

Balanced speed, cost, and intelligence.A powerful and efficient next-generation model for a wide range of professional and creative tasks. It delivers high-quality results in complex reasoning, coding, and large-context processing.

Google's most capable Flash model for agentic workflows and multimodal reasoning.

Google's fast and affordable workhorse model, tuned for coding, long multi-step tasks and autonomous agents, with adjustable reasoning effort to balance quality, cost and speed. With a context window of 1 million tokens, it handles large documents and extensive projects.

Google's flagship reasoning model and the first of the Gemini 4 family: complex software engineering, mathematics and science, autonomous multi-step agents and professional work in finance and law. Reads text, images, video, audio and PDF, with a context window of 1M tokens.

Default flash image model. Best balance of speed, cost and quality. Supports thinking budget and Google Search grounding (incl. image search subtype).

Pro-tier image model for advanced asset production. Strong instruction following, internal thinking always on. Supports Google Search grounding.

Google's highest quality video model: 4-8 s videos with native sound, up to 4k, from text, a starting image, first and last frame or up to 3 reference images. Price per 1M seconds of 720p video.

Faster and cheaper Veo 3.1: 4-8 s videos with native sound, up to 4k, from text, images or up to 3 reference images. Price per 1M seconds of 720p video.

The cheapest Veo: 4-8 s videos with native sound in 720p or 1080p from text or a starting image (no reference images). Price per 1M seconds of 720p video.

DeepSeek's latest lightweight model (V4.1 Flash), built for agentic tasks with an excellent balance of performance and cost. It understands text and images and outperforms the larger DeepSeek V4 Pro on most benchmarks while staying fast and affordable.

A fast and very affordable model from the DeepSeek V4 generation, optimized for efficiency in everyday tasks, automation and high request volumes. With a context window of 1 million tokens, it also copes with long documents.

The large open-weight model of the DeepSeek V4 generation, designed for demanding reasoning, coding and agentic workflows. With a context window of 1 million tokens, it handles extensive documents and large projects at a very low price.

A specialized coding model with strong instruction following and reliable performance across long contexts. It is designed for long-horizon development, agentic coding, tool use, and complex problem-solving.

A specialized coding model with strong instruction following and reliable performance across long contexts. It is designed for long-horizon development, agentic coding, tool use, and complex problem-solving.

Open-source multimodální model zaměřený na programování, vizuální porozumění a agentní úlohy. Dokáže pracovat s textem, obrázky i videem a převádět vizuální návrhy na funkční kód.

Moonshot AI's flagship open-weight model of the K3.1 series with a 1M-token context window. Strong at coding, agentic tasks and tool use; understands text, images and video and always reasons before answering.

The fast variant of Kimi K3.1 Pro: the same model and 1M-token context, answering 2–3× faster at twice the price. Suited to interactive coding and agentic tasks where latency matters.

xAI’s advanced flagship model, designed for accurate responses, reliable instruction following, and effective tool use. With a context window of up to 1 million tokens, it is well suited for complex analysis, agentic workflows, and processing extensive documents.

SpaceXAi flagship model that was built for long-horizon agentic coding and real-world knowledge work

The latest frontier Grok model for coding, agentic tasks and professional knowledge work, noticeably stronger than Grok 4.6 at the same price and speed. With a 500k-token context window, it handles large codebases, long documents and multi-step workflows.