DeepSeek V4.1 Flash is now available in Token Pass
By Steven Van ·
The $30 per month Pro team subscription includes $80 of credit, and DeepSeek V4.1 Flash costs $0.15 per million input tokens.
DeepSeek V4.1 Flash is now a featured model in Vercel's Token Pass, joining Kimi K3 and GLM 5.3. Vercel says each model is accessed through its preferred provider and that more models are coming.
What the model uses
The model ID is deepseek/deepseek-v4.1-flash. It always uses DeepSeek's latest V4.1 Flash weights, currently the 10/17 snapshot, following DeepSeek's own naming convention. In AI Gateway, requests show Vercel Token Pass (token-pass) as the provider, which currently serves the models through Wafer.
Credit is deducted at list price. For DeepSeek V4.1 Flash that is $0.15 per million input tokens, $0.003 for cached input and $0.60 for output.
How Token Pass works
Token Pass is a subscription for Pro teams that costs $30 per team each month and grants $80 of credit per billing period. It runs through AI Gateway and works with any coding agent that connects with an API key, such as fx, Pi, Cline and Command Code, or with your own app built on the AI SDK. Zero data retention is on by default, and prompts are not used for training.
To add the model, use the Token Pass tab in the team's AI Gateway dashboard. A Token Pass API key only works with the featured models. Credit is shared by the whole team and does not roll over. When it runs out, Token Pass requests fail until the next billing period, with no extra charge.
Availability
Token Pass is in Early Access with a limited number of spots, and a waitlist appears when it is full. Hobby teams need to upgrade to Pro first, and Enterprise teams can't subscribe. Setup steps are in the Token Pass documentation.