Inception announced Mercury 3 on Oct. 19, 2026, offering businesses another language-model option for applications where response time matters. Its launch prices and distribution through outside platforms give developers practical details to evaluate alongside the company’s performance claims.
This Rapid News Brief examines the announced offering, not an independently tested speed ranking.
Availability and pricing
Inception says Mercury 3 is available through its own API and through OpenRouter, Baseten and Vercel. The company lists a context window of 1 million tokens and says the service supports an API format compatible with OpenAI’s.
The announcement gives standard prices of 20 cents per million input tokens and 70 cents per million output tokens, with a 50% launch discount. Those discounted rates are 10 cents and 35 cents, respectively.
OpenRouter’s catalog displays the same discounted prices and context size. It also lists support for tool calling and structured outputs, features relevant to applications that connect a model to business systems or require consistently formatted responses.
Routing is part of the decision
The distributor says Mercury 3 is not currently available through its EU in-region routing. That is a limitation of the listed route, not evidence that the model is unavailable everywhere in Europe. Teams with location-specific deployment requirements should distinguish model access from the routing options offered by a particular distributor.
OpenRouter labels the model’s release date Oct. 16. Inception’s public announcement is dated Oct. 19; neither date makes this an Oct. 20 product launch.
Inception promotes Mercury 3 as a diffusion-based model aimed at latency-sensitive work, including customer support and browser agents. Its claims of superior speed and quality come from the company and should not be treated as universal outcomes.
For buyers, the operational test is whether a model completes their tasks accurately within the required response time and deployment constraints. A token price or headline generation rate alone does not establish the cost of a successful workflow.