DeepSeek
5 models · est. 2023
Open-weight frontier reasoning at disruptive economics
Providers
The labs behind the models.
Nineteen providers across the Chinese and global ecosystems, profiled, compared and kept current.
Showing 20 of 20 providers
Chinese ecosystem
Ten labs setting the pace
Featured first across Webparam, the fastest-moving ecosystem in AI.
Alibaba Qwen
8 models · est. 2023
The broadest open model family in the world
Moonshot AI
5 models · est. 2023
Long-context agentic intelligence (Kimi)
MiniMax
6 models · est. 2021
Multimodal breadth: text, video, speech
Tencent Hunyuan
3 models · est. 2023
Ecosystem-scale multimodal models
ByteDance Doubao
4 models · est. 2023
Consumer-scale LLMs and generative video
Baidu ERNIE
2 models · est. 2019
China's longest-running LLM lineage
Zhipu AI
5 models · est. 2019
Open agentic GLM family
01.AI
1 models · est. 2023
Efficient bilingual frontier models
StepFun
2 models · est. 2023
Multimodal foundation models
Global
The global frontier
The reference points: frontier capability, open weights at scale and the enterprise stacks behind them.
Webparam
20 models · est. 2026
One account, one wallet, one API key — billed in rand
OpenAI
4 models · est. 2015
The reference point for frontier capability
Anthropic
3 models · est. 2021
Frontier models with a safety-first identity
Google
5 models · est. 1998
Frontier multimodality at planetary scale
Meta
2 models · est. 2004
Open weights at the frontier (Llama)
Mistral
4 models · est. 2023
Europe's frontier lab; efficient and open
xAI
2 models · est. 2023
Real-time-grounded frontier models
Cohere
3 models · est. 2019
Enterprise retrieval and language stack
Amazon
2 models · est. 1994
Price-performance models across modalities
NVIDIA
1 models · est. 1993
Open reasoning models tuned for its stack
