# AI Models & Tools

> Three different lenses: API-routed token usage, website visits and capability benchmark results.

Canonical: https://ainextia.com/ai-models/
Publisher: https://ainextia.com/

## Data-source scopes
- OpenRouter: OpenRouter API traffic; token volume by individual model variant. As-of date in fallback snapshot: 2026-11-05. See the HTML page for any updated API snapshot: https://ainextia.com/ai-models/
- Similarweb: consumer AI website traffic; not model usage.
- Artificial Analysis: specified model benchmark configurations; not general usage.

## Official model providers
- [OpenAI](https://openai.com/news/) — ChatGPT and GPT models
- [Anthropic](https://www.anthropic.com/news) — Claude and coding agents
- [Google](https://deepmind.google/models/) — Gemini and DeepMind research
- [DeepSeek](https://www.deepseek.com/) — Open and hosted language models
- [Z.ai](https://z.ai/) — GLM language models
- [Qwen](https://qwenlm.github.io/) — Alibaba's Qwen model family
- [Moonshot AI](https://www.kimi.com/) — Kimi models and assistants
- [MiniMax](https://www.minimax.io/) — Multimodal and language models
- [Tencent](https://hunyuan.tencent.com/) — Hunyuan model family
- [Xiaomi](https://github.com/XiaomiMiMo) — MiMo model family
- [NVIDIA](https://developer.nvidia.com/) — Nemotron models and inference
- [Meta](https://ai.meta.com/) — Open-weight model research
- [xAI](https://x.ai/) — Grok models and products
- [StepFun](https://stepfun.ai/) — Step model family
