# Vercel AI Gateway adds DeepSeek V4.1 Flash 1017 Ultrafast

published: 2026-11-07T20:05:26.600Z
topic: infrastructure
source: https://vercel.com/changelog/deepseek-v4-1-flash-1017-ultrafast-now-available-on-ai-gateway
url: https://aiagentnews.io/stories/vercel-ai-gateway-adds-deepseek-v4-1-flash-1017-ultrafast

Vercel has made DeepSeek V4.1 Flash 1017 Ultrafast available on AI Gateway, served through Fireworks. The model can generate more than 1,000 tokens per second and is in public preview.

Vercel says DeepSeek V4.1 Flash 1017 Ultrafast is now available on AI Gateway, with the model served through Fireworks. It runs the October 17 update of DeepSeek V4.1 Flash on Fireworks' accelerated serverless endpoint, and Vercel says it can generate upwards of 1,000 tokens per second.

The model has a 1M-token context window for large codebases and long inputs. It is in public preview with no waitroom or session time limit, and Vercel positions it for interactive coding, reasoning and agentic workflows.

Developers can select it by setting deepseek/deepseek-v4.1-flash-1017-ultrafast as the model name in the AI SDK. For Claude Code, Codex, Cursor and similar tools, Vercel says to install the latest Vercel CLI and run npx vercel ai-gateway setup, then pick the model in the agent. AI Gateway says it applies provider pricing with no markup and charges no platform fee on inference, including Bring Your Own Key requests.

## Why it matters if you build agents

Agent builders can try a fast 1M-token DeepSeek model by changing a single model name behind the AI Gateway endpoint they may already call. Because it is in public preview with no session time limit, it is practical to test long-running agent loops now. The source does not state the model's own per-token price, so check the provider pricing before scaling up.