Inception Labs

Diffusion language model startup behind Mercury

inceptionlabs.aiPalo Alto, United States

Inception Labs is a Palo Alto startup founded in 2024 that builds diffusion-based large language models. Its Mercury family generates many tokens in parallel for fast code and text generation.

Catalogue

All models by Inception Labs

1 models · 1–1

Select models to compare	Model	Released	Context	Input /M	Output /M	Parameters	Type	I/O	Openness
	Mercury Edit 2Inception Labs	March 30, 2026	32K	—	—	—	Text	text → text	Closed

Page 1 of 1

Previous Next