DeepSeek V4 Flash now runs updated weights on AI Gateway
- Source
- vercel.com
- Author
- Jerilyn Zheng
- Date

Context
Keeping a model identifier unchanged does not necessarily keep the underlying model unchanged. Vercel's July 31 AI Gateway update says requests using deepseek/deepseek-v4-flash automatically receive updated weights, without an application code change.
At the announcement, DeepSeek was the only provider serving those weights. Additional providers, including providers offering zero data retention, were described as forthcoming. The update therefore changed the default model behind an existing identifier, while provider availability remained a separate consideration.
“DeepSeek V4 Flash now runs on updated weights by default on AI Gateway, with notably stronger agentic capabilities. On Terminal-Bench, it scores 82.7, up 25.8 points from 56.9 in the April preview.”
“Requests to deepseek/deepseek-v4-flash pick up the new weights automatically, with no change to the model ID or your code.”
“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.”
Checking sign-in…
Loading comments…






