Providers
19 AI labs, one API key. Browse models by provider — every major Chinese and Western lab in one place.
OpenAI makes the GPT series, the most widely-used AI models.
Alibaba Qwen (Tongyi Qianwen) is one of China's most capable model families.
Zhipu AI (BigModel) is the maker of the GLM series, including the flagship GLM-5.
Anthropic makes the Claude series, widely regarded as top-tier for coding, careful reasoning, and long-context analysis.
ByteDance Doubao (Seed) models are widely used inside ByteDance products.
Google Gemini models are natively multimodal — handling text, image, video, and audio in one model.
DeepSeek is a leading Chinese AI lab known for high-quality open-weight reasoning models at extremely low cost.
MiniMax produces the M-series multimodal models, strong at creative writing, voice, and content generation.
Xiaomi MiMo models are lightweight and affordable, optimized for mobile and edge scenarios with strong Chinese support.
Moonshot AI makes the Kimi series, known for excellent long-document comprehension and Chinese-language quality.
Baidu ERNIE (Wenxin) is one of China's earliest LLM families.
xAI makes the Grok series.
StepFun makes the Step series of models, focused on cost-efficient inference for high-volume applications.
Shanghai AI Lab (OpenLabSSS) develops the InternLM/InternVL series, strong open-source models for research and vision tasks.
Tencent Hunyuan models power Tencent's products.
Kuaishou Kat models are specialized for code generation, from the short-video platform Kuaishou.
Agnes AI AI models.
Meta's Llama series is the most popular open-weight model family.
Meituan LongCat is Meituan's in-house model, used across its food-delivery and local-services products.
xAI AI models.