# Mercury 3 from Inception now available on Vercel AI Gateway

published: 2026-10-19T22:05:25.947Z
topic: infrastructure
source: https://vercel.com/changelog/mercury-3-now-available-on-ai-gateway
url: https://aiagentnews.io/stories/mercury-3-from-inception-now-available-on-vercel-ai-gateway

Vercel has added Mercury 3, a model from Inception, to AI Gateway with 50% off launch pricing. Inception reports strong benchmark results and generation speeds above 1,000 tokens per second.

Mercury 3 is now available on AI Gateway, and Vercel is offering 50% off its launch pricing. Inception says the model posts strong results across a range of benchmarks and generates text above 1,000 tokens per second. It uses diffusion to generate text through parallel refinement.

Inception positions Mercury 3 for latency-sensitive workloads such as browser agents, real-time search, interactive education and customer support. The model has a 1M-token context window for long documents and extended conversation histories. Developers set AI_GATEWAY_API_KEY and use inception/mercury-3 as the model name in AI SDK calls.

The changelog also says the model can be selected in Claude Code, Codex and Cursor after installing the latest Vercel CLI and running npx vercel ai-gateway setup. AI Gateway provides a unified API, usage and cost tracking, retries, failover and budgets for API keys. It reflects provider pricing with no markup and charges no platform fee on inference, including Bring Your Own Key requests.

## Why it matters if you build agents

Agent builders can call a fast-generating model with a 1M-token context window through the same gateway key and model setup they use for other providers. Teams that route latency-sensitive steps, such as browser or search agents, can test it without a separate integration.

The source does not say how long the 50% launch discount lasts, so check current pricing before budgeting around it.