DeepSeek V4 Flash now runs updated weights on AI Gateway
Quick Answer
DeepSeek V4 Flash now operates with updated weights on AI Gateway, achieving a score of 82.7 on Terminal-Bench, a significant increase of 25.8 points.
Quick Take
Users can access these new weights automatically without changing the model ID or code, with other providers expected to follow next week.
Key Points
- DeepSeek V4 Flash scores 82.7 on , up from 56.9 in April.
- Automatic access to updated weights with no code changes required.
- DeepSeek is currently the only provider with updated weights.
- AI Gateway charges no platform fee on inference requests.
- Users can set the model to deepseek/deepseek-v4-flash in the AI SDK.
Source Excerpt
DeepSeek V4 Flash now runs on updated weights on AI Gateway, served by DeepSeek by default, with stronger agentic capabilities.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from Vercel AI
See more →
The Agent Stack
The Agent Stack by Vercel AI provides essential building blocks for creating production-grade agents, enabling seamless integration across multiple AI models and secure operations. It features components like AI Gateway for model routing, Workflow SDK for durable execution, and Vercel Connect for scoped access, streamlining agent development and deployment across various platforms.
