什么是 AI Gateway？

Cloudflare 推出 AI Gateway 花费控制功能，帮助企业管理 AI 使用成本并实现精细化预算分配。

DeepSeek 在 2026 年 5 月迅速增长至 AI Gateway 的第三大模型，但其花费占比仍低于 1%，Anthropic 仍主导高价值使用场景。

Vercel 推出 AI Gateway API 密钥预算功能，可设置消费上限以控制 AI 成本。

Claude Fable 5 现已通过 AI Gateway 提供，其在多步骤任务处理和代码审查方面表现更优。

Vercel AI Gateway now supports sorting model providers by cost, latency, or throughput to help developers choose services more efficiently.

Vercel's Chat SDK now includes support for Facebook Messenger via a dedicated adapter, enabling unified chat deployment across platforms.

Cloudflare extends its cloud access security broker (CASB) to support Anthropic's Claude Compliance API, allowing security teams to monitor Claude usage directly in the Cloudflare dashboard without endpoint agents. This addresses the issue of AI application security governance lagging behind adoption, providing full lifecycle security coverage from API calls to data processing.

文章探讨了集成开发环境（IDE）作为AI质量变量的重要性，强调在AI议程中考虑IDE的选择，以优化AI工具的性能和开发者效率。通过分析AI门控器的管理作用与局限性，指出IDE提供的代码库结构知识能显著影响AI辅助开发的效果，特别是在AI生成代码日益增多的背景下。

Vercel integrates xAI's Grok Imagine Video 1.5 into AI Gateway, enabling unified model access and improved video generation quality and efficiency.

Vercel's AI Gateway plugin enables 42% of websites to integrate multimodal AI models via a single API key, supporting text, image, video, and audio across all providers.

Vercel's AI Gateway now supports the fast mode of Opus 4.7, significantly improving audio processing speed.

Vercel AI Gateway 现已支持 Grok 4.3，该模型更新了2025年12月的知识库截止点与100万字符的上下文窗口，增强了准确性、工具调用与指令遵循能力。开发者可通过AI SDK使用此模型。

Vercel has integrated Gemini 3.5 Flash into its AI Gateway, offering developers low-latency, cost-efficient access to Google's latest lightweight model, though no performance benchmarks or pricing details are disclosed.

Vercel AI Gateway 每月平均恢复超过 1T tokens，通过零标记策略和智能重试机制实现。

Vercel CLI 现在支持创建 AI Gateway API 密钥并设置预算，但信息密度较低，缺乏深度技术细节。

Data from Vercel's AI Gateway reveals Google leads in production scale, Anthropic dominates coding and spend, OpenAI grows rapidly post-5.4, and OSS models gain traction.

文章内容信息密度低，缺乏具体技术细节和深度分析，仅提及DeepSeek进入AI领域竞争，未提供实质性技术或行业洞察。