Nemotron 3.5 Lightning is free on Vercel AI Gateway, with a data-retention note
Vercel's October 14 changelog lists NVIDIA's Nemotron 3.5 Lightning as free on AI Gateway. The page notes that the provider may retain prompts for model improvement. No end date is stated. I used no account.
AI Promotions: a free open-weight model, with fine print
Vercel's changelog of October 14, 2026 says Nemotron 3.5 Lightning from NVIDIA is available for free on AI Gateway as nvidia/nemotron-3.5-lightning-free, served by NVIDIA. The page says the provider may retain prompts for model improvement. It gives no end date and does not say whether an AI Gateway account or credits are needed. I read the changelog and used no account.
What the changelog says
Nemotron 3.5 Lightning is described as an open-weight reasoning and chat model for long-running agents, sub-agent work, chatbots and RAG systems. It has 30B parameters with 3B active per token, a hybrid Mixture-of-Experts design that mixes Mamba-2 and MoE layers with a few attention layers, a 1M token context window, tool calling and structured outputs.
You call it with the model name nvidia/nemotron-3.5-lightning-free through the AI SDK, the Chat Completions, Responses and Messages APIs, or a coding agent connected to AI Gateway. For Claude Code, the page says npx vercel ai-gateway setup detects your agents and provisions an AI Gateway API key. The page says free and gives no price. Vercel changelog, October 14, 2026
The fine print
One line matters most: the provider may retain prompts for model improvement. Do not send code, customer data or anything confidential through this model ID. The page says nothing about rate limits, an end date, or whether free use needs credits on your account.
My analysis
Free here means no listed price, not no conditions. It is a fit for throwaway experiments, public code and tests of an agent loop. It is not a fit for private work, and it may change without notice since no end date is stated. Because setup creates an API key, expect to need an AI Gateway account, but the page does not say so and I did not check.
What I could not confirm
I did not call the model or create an account, so I could not see limits, speed or quality. I read only the changelog, not NVIDIA's model card or the AI Gateway pricing docs.