The Pulse
Vercel's AI Gateway Adds DeepSeek V4.1 Flash 1017 Ultrafast at Upwards of 1,000 Tokens per Second
DeepSeek V4.1 Flash 1017 Ultrafast is now available on AI Gateway through Fireworks and can generate upwards of 1,000 tokens per second. The model has a 1M-token context window and is available in public preview.
AI.info Team · Written with AI · fact-checked ·
Ultrafast serves the October 17 update of DeepSeek V4.1 Flash through Fireworks' accelerated serverless endpoint
DeepSeek V4.1 Flash 1017 Ultrafast is now available on AI Gateway. The model is suited to interactive coding, reasoning and agentic workflows where applications need to generate output and move to the next step quickly.
The model is available in public preview without a waitroom or session time limit
The model can generate upwards of 1,000 tokens per second. Its 1M-token context window is for working with large codebases and long inputs.
AI SDK and coding agents can use the model
Users set deepseek/deepseek-v4.1-flash-1017-ultrafast as the model name. An AI SDK example calls streamText with a prompt for a TypeScript helper that retries a failed request with backoff. For Claude Code, Codex, Cursor and more, users install the latest Vercel CLI and run npx vercel ai-gateway setup. After setup, users select the model in the agent.
AI Gateway provides a unified API and built-in controls
AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover and performance optimizations for higher-than-provider uptime. The service includes built-in custom reporting, Zero Data Retention support, budgets for API keys, routing rules and more. AI Gateway reflects provider pricing with no markup and does not charge a platform fee on inference, including on Bring Your Own Key (BYOK) requests.