Site-wide model, accessible with a single key

Claude Fable 5 focuses on long-form writing and narrative expression, making it ideal for generating stories, copy, and creative content.

Claude Fable 5.1 focuses on long-form writing and narrative expression, making it ideal for generating stories, copy, and creative content.

Claude Haiku 4.5 emphasizes low latency and low cost, making it suitable for high-concurrency classification, entity extraction, and lightweight question-answering tasks.

Claude Haiku 5.5 emphasizes low latency and low cost, making it suitable for high-concurrency classification, entity extraction, and lightweight question-answering tasks.

Claude Opus 4.5 is Claude's high-end model, suitable for complex reasoning, code architecture, and challenging analytical tasks.

Claude Opus 4.6 is Claude's high-end model, suitable for complex reasoning, code architecture, and challenging analytical tasks.

Claude Opus 4.7 is Claude's high-end model, suitable for complex reasoning, code architecture, and challenging analytical tasks.

Claude Opus 4.8 is Claude's high-end model, suitable for complex reasoning, code architecture, and challenging analytical tasks.

Claude Opus 5 is Claude's high-end model, suitable for complex reasoning, code architecture, and challenging analytical tasks.

Claude Opus 5.5 is Claude's high-end model, suitable for complex reasoning, code architecture, and challenging analytical tasks.

Claude Sonnet 4.5 strikes a balance between capability and cost, making it suitable for everyday coding, writing, and general-purpose agent workflows.

Claude Sonnet 4.6 strikes a balance between capability and cost, making it suitable for everyday coding, writing, and general agent workflows.

Claude Sonnet 5 strikes a balance between performance and cost, making it suitable for everyday coding, writing, and general-purpose agent workflows.

Claude Sonnet 5.5 strikes a balance between capability and cost, making it suitable for everyday coding, writing, and general agent workflows.

Gemini 2.5 Flash is designed for high-throughput real-time scenarios and is suitable for content processing, conversations, and large-scale calls.

Gemini 2.5 Flash Lite is a lightweight version of Gemini, suitable for simple Q&A and batch processing tasks that require high throughput and low latency.

Gemini 2.5 Pro is Gemini's flagship model, suitable for complex reasoning, long-document comprehension, and multi-step tasks.

Gemini 3 Flash Preview is designed for high-throughput real-time scenarios and is ideal for content processing, conversations, and large-scale calls.

Gemini 3.1 Flash Lite is a lightweight version of Gemini, suitable for simple question-and-answer and batch processing tasks that require high throughput and low latency.

Gemini 3.1 Flash Lite Preview is a lightweight version of Gemini, suitable for simple Q&A and batch processing tasks that require high throughput and low latency.

Gemini 3.1 Pro Preview is Gemini's flagship model, suitable for complex reasoning, long-document comprehension, and multi-step tasks.

Gemini 3.5 Flash is designed for high-throughput real-time scenarios and is ideal for content processing, conversations, and large-scale calls.

Gemini 3.5 Flash Lite is a lightweight version of Gemini, suitable for simple question-and-answer and batch processing tasks that require high throughput and low latency.

Gemini 3.6 Flash is designed for high-throughput real-time scenarios and is suitable for content processing, conversations, and large-scale calls.

Gemini 3.7 Flash is designed for high-throughput real-time scenarios and is suitable for content processing, conversations, and large-scale calls.

Gemini 3.8 Flash is designed for high-throughput real-time scenarios and is suitable for content processing, conversations, and large-scale calls.

Gemini 4 Argon is part of the Gemini 4 series and is designed for high-difficulty reasoning and complex task processing.

The GLM-4.5-Air is the lightweight version of the GLM-4 series, suitable for cost-sensitive, high-frequency calls.

GLM-4.6 is part of the Zhipu GLM-4 series and is suitable for Chinese conversation, writing, and tool invocation.

GLM-4.6V is GLM's visual understanding model, suitable for image-based question-answering and image-text hybrid tasks.

GLM-4.7 is part of the Zhipu GLM-4 series and is suitable for Chinese conversation, writing, and tool invocation.

The GLM-5 is part of the Zhipu GLM-5 series and is designed for complex reasoning, coding, and agent-based scenarios.

The GLM-5-Turbo is the high-performance version of the GLM-5 series, suitable for general-purpose tasks where response speed is critical.

The GLM-5.1 is part of the Zhipu GLM-5 series and is designed for complex reasoning, coding, and agent applications.

The GLM-5.2 is part of the Zhipu GLM-5 series and is designed for complex reasoning, coding, and agent applications.

The GLM-5.3 is part of the Zhipu GLM-5 series and is designed for complex reasoning, coding, and agent applications.

GLM-5.3-Flash is the lightweight version of the GLM-5 series, suitable for cost-sensitive applications involving frequent calls and simple tasks.

GLM-5.3-FlashX is the high-speed, lightweight version of the GLM-5 series, suitable for high-concurrency, low-latency batch calls.

GPT-5.5 is part of OpenAI's GPT-5 series and is suitable for complex reasoning, coding, and production-level general-purpose tasks.

GPT-5.6 Luna is part of OpenAI's GPT-5 series and is suitable for complex reasoning, coding, and production-level general-purpose tasks.

GPT-5.6 Sol is part of OpenAI's GPT-5 series and is suitable for complex reasoning, coding, and production-level general-purpose tasks.

GPT-5.6 Terra is part of OpenAI's GPT-5 series and is suitable for complex reasoning, coding, and production-level general-purpose tasks.

GPT-6 Astra is part of OpenAI's GPT-6 series and is well-suited for complex reasoning, coding, and multi-step agent tasks.

GPT-6 Sol is part of OpenAI's GPT-6 series and is suitable for complex reasoning, coding, and multi-step agent tasks.

GPT-6.1 Sol is part of OpenAI's GPT-6 series and is suitable for complex reasoning, coding, and multi-step agent tasks.

Kimi K2.5 is part of Moonshot's Kimi series and is suitable for long-form text processing, conversation, and content generation.

Kimi K2.6 is part of Moonshot's Kimi series and is suitable for long-form text processing, dialogue, and content generation.

Kimi K2.7 Code is designed for programming scenarios and is suitable for code generation, refactoring, and repository-level understanding.

Kimi K3 is Moonshot's next-generation Kimi model, designed for reading long documents, organizing information, and handling complex conversations.