AI Agent StoreDeepSeek · Hosted serving variant
DeepSeek V4.1 Flash Fast
AI Gateway offers this version of V4.1 Flash for faster serving. We do not yet have comparable speed measurements, and results for the standard service are not assigned to this version.
Provider documentation discovered 2026-10-03 · Checked 2026-10-11. The provider launch date is unverified; discovery is a separate date.
We still need comparable independent tests. You can compare the confirmed prices and limits below, but we cannot yet give this version an Agent Fit or Practical Value rating, or recommend a thinking setting. Try it on your own tasks before switching. Results for an older or differently configured version may not apply.
What you can use today
Token prices measure the small pieces of text the model reads or produces. Capacity, often called the context window, is the amount it can handle in a request. A dash means we could not confirm the price.
Exact AI Gateway route ID: deepseek/deepseek-v4.1-flash-fast. The exact Gateway API declares text/image input, tool use, reasoning controls, a 1,048,576-token context and 1,000,000 output tokens. Native mappings and independent variant measurements are pending. Test accuracy, speed and reliability on your own tasks.
Thinking settings listed by AI Gateway: none, low, medium, high. These controls apply to this AI Gateway service. We have not confirmed how they match the model maker’s settings, which is the default, or how they perform in comparable independent tests. Verify the exact Gateway catalog entry ↗
The route is absent from the October 9 Gateway catalog and its model page returns 404; verify access before planning new work. Historical price and capability observations below retain their original scope.
Limits belong to this hosted route and may differ from another deployment of V4.1 Flash.
Gateway rates observed October 4 are $0.30 input, $1.20 output and $0.006 cache read per million tokens. These are routed quotes, separate from direct-provider price schedules. This is a historical offer, not a currently verified quote. Your bill can also depend on discounts for reused input, which provider handles the request, speed options, tools, retries and charged thinking tokens.
Check availability and plans ↗
Test it before switching
Give each model the same examples, tools and definition of success. Record correct outcomes, response time, retries, token use and total cost. Even at the same price per token, a different version can cost more if it uses more tokens.
Compare caching and long-context costs ↗