[Feature] Integration with models.dev for auto-updating AI Model metadata & pricing (Zero-Maintenance Architecture)

1Panel Version

v2.2.5

Please describe your needs or suggestions for improvements

1. Feature Description / 功能描述

• Relates to / 关联 Issue: #13036 (Thanks to @bagchi-akash for initiating the discussion on expanding model providers).

Inspired by the demand in #13036 to expand model providers (such as OpenRouter, NVIDIA, Nous, etc.), manually hardcoding and updating provider endpoints, model context windows, capabilities, and token pricing for each LLM vendor is unsustainable.

We propose leveraging models.dev (an open-source, community-driven metadata standard) as the unified data source for 1Panel's AI provider configurations.

受 #13036 中关于扩展大模型服务商（如 OpenRouter, NVIDIA, Nous 等）需求的启发，考虑到手动在代码中硬编码和维护各大厂商模型的上下文长度、功能支持及 Token 价格对维护者来说成本极高且难以持续。
建议引入开源且由社区共同维护的 models.dev 作为 1Panel 统一的 AI 模型元数据标准数据源。

2. Why this is the Optimal Architecture / 为什么这是最佳架构方案？

• Zero Maintenance for 1Panel Team (官方零维护成本):
  LLM vendors (e.g., DeepSeek, OpenAI, Huawei Cloud, NVIDIA) and global open-source contributors actively update models.dev whenever new models or price adjustments occur. 1Panel simply consumes the upstream standard JSON data without needing manual PRs or hotfixes for every model change.
  各大大模型厂商（如 DeepSeek、OpenAI、华为云、NVIDIA 等）及开源社区会在模型更新或调整价格的第一时间向 models.dev 提交 PR。1Panel 只需消费上游的标准 JSON 数据，无需在每次模型变动时手动发版或改代码。

• Decoupled Architecture (架构解耦):
  1Panel only needs to render the UI and route requests based on standard JSON fields. The core code remains stable regardless of how many providers or models are added upstream.
  1Panel 仅需根据标准 JSON 字段渲染前端与调度逻辑，架构高度解耦，无论上游增加多少模型，底层代码无需频繁修改。

3. References / 参考资料

• Upstream Project: https://github.com/anomalyco/models.dev
• Standard JSON Endpoint: https://models.dev/api.json

Please describe the solution you suggest

Future Version Update.

Additional Information

No response