Summary
Vercel added DeepSeek V4 Pro and DeepSeek V4 Flash to AI Gateway on April 23, 2026, making both models available through the AI SDK with a 1M-token context window as the default. Vercel highlights the models for agentic coding, mathematical reasoning, long-horizon workflows, and high-volume lower-cost tasks depending on the variant.
What changed
Vercel AI Gateway added DeepSeek V4 Pro and DeepSeek V4 Flash with default 1M-token context support.
Why it matters
This broadens Vercel's model layer beyond frontier US vendors and makes long-context, lower-cost reasoning models easier to operationalize inside existing app stacks. It also reinforces the idea that the gateway vendor, not the model vendor, increasingly owns routing, observability, and deployment ergonomics.
Evidence excerpt
Vercel said DeepSeek V4 Pro and V4 Flash are now available on AI Gateway with a 1M token context window by default and can be called through the AI SDK as deepseek/deepseek-v4-pro or deepseek/deepseek-v4-flash.