DeepSeek V4 Flash now runs updated weights on AI Gateway
PreviewVerifiedAdded Sep 22, 2026
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.
Topics: AI agents, Pricing, Developer tools, Coding agents
Summaries of vendors' own notes. Product names and logos belong to their owners; logos via logo.dev.
More AI SDK and AI Gateway releases
Every AI SDK and AI Gateway release| Date | Release | Type |
|---|---|---|
| Jul 31 | AI Gateway now supports team and project spend budgets Update | Update |
| Jul 31 | AI Gateway logs now have a dedicated page Update | Update |
| Jul 31 | Expanded search for workflow runs in Vercel Observability Preview | Preview |
| Jul 31 | 10x more capacity for Laguna S 2.1 on AI Gateway Update | Update |
| Jul 31 | AI Gateway adds unified fast mode support Beta | Beta |
| Jul 30 | Run multiple isolated agents in a single Sandbox Update | Update |
| Jul 30 | MiniMax H3 now available on AI Gateway Update | Update |
| Jul 30 | Inkling Small from Thinking Machines is now available on AI Gateway Update | Update |
