# Inception Launches Mercury 3 With One-Million-Token Context Window

By Simon Yoon

Canonical URL: https://www.tokenpost.com/news/technology/35427
Published: 2026-10-20T07:46:36.000Z
Updated: 2026-10-20T07:46:36.000Z
Section: Technology

> Standard pricing is $0.20 per million input tokens and $0.70 per million output tokens, with introductory rates set at half those amounts.

Inception launched Mercury 3 on Oct. 19, giving developers a diffusion-based language model with a one-million-token context window and access through four distribution channels.

A context window determines how much text a model can consider in a request. Mercury 3's standard price is $0.20 per million input tokens and $0.70 per million output tokens. Introductory rates cut those prices in half, to $0.10 and $0.35, respectively. Input tokens are text sent to the model; output tokens are text it generates.

In company-run release tests, Mercury 3 generated 1,068 tokens per second on NVIDIA GPUs. It also scored ahead of GPT-6 Luna and Claude Haiku 5.5 on multiple benchmarks in those tests. The comparisons describe the release tests rather than performance on every task.

Developers can access Mercury 3 through Inception's API, OpenRouter, Baseten and Vercel. The model is compatible with the OpenAI API, allowing applications built around that interface to use it.