Vercel AI Updates: October 20, 2026

1. Inception’s Mercury 3 Diffusion Model Arrives on AI Gateway at 50% Off

Vercel. AI Gateway now serves Mercury 3 from Inception (inception/mercury-3) with 50% off launch pricing. Mercury 3 generates text with diffusion, refining output in parallel rather than token by token, and Inception reports generation speeds above 1,000 tokens per second. It has a 1M-token context window and targets latency-sensitive workloads such as browser agents, real-time search, and customer support. The model can be selected in Claude Code, Codex, Cursor, and other coding agents after running setup with the latest Vercel CLI. Source

2. MiniMax M3.1 Flash Is Free on AI Gateway Until October 22

Vercel. AI Gateway added MiniMax M3.1 Flash (minimax/minimax-m3.1-flash), an open-weight multimodal model for coding and agentic work with a 1M-token context window, free until October 22. MiniMax reports improvements over its earlier models on complex agentic tasks, computer use across applications, coding (especially frontend work, targeted bug fixes, and component tests), and cybersecurity. The free rate applies only when MiniMax itself serves the request, so developers need to pin the provider with order in the gateway provider options; pinned requests bill at MiniMax’s standard rate after the offer ends. It is available through the AI SDK and the Chat Completions, Responses, and Anthropic Messages APIs. Source