Inception
Inception develops the Mercury family of diffusion-based large language models for production applications. · Model developers
How they use Nebius AI Cloud
Inception lists Nebius in its subprocessor register as a cloud infrastructure provider used to support the training and serving of large language models.
Inception develops Mercury, its proprietary family of diffusion-based large language models. It provides enterprise customers with access through cloud-based APIs and enterprise deployments.
Sources
1 sourceCompany context
inceptionlabs.aiInception, operated by Inception AI, Inc., develops the Mercury family of diffusion-based large language models for developers and enterprise teams. Its models generate and refine text in parallel for applications such as coding assistants, automated business workflows, search and conversational agents. Its products include the flagship Mercury 3 model and Mercury Voice for voice applications. Developers access Mercury through Inception’s API platform and partner platforms, while enterprise offerings provide dedicated capacity, autoscaling, compliance controls and configurable data retention.
InceptionRecent updates
5 updatesInception launches Mercury 3
Mercury 3 became available through Inception's API, OpenRouter, Baseten and Vercel. The model supports a 1 million-token context window and delivers 1,068 tokens per second on NVIDIA GPUs, according to Inception.
InceptionInception releases Mercury Decide
AlphaSignal reported that Inception released Mercury Decide through free early access on OpenRouter. The structured decision model returns choices, scores or yes/no answers with confidence probabilities for routing, classification and evaluation workflows.
AlphaSignalMercury Voice becomes generally available to enterprises
Inception made Mercury Voice available to enterprise customers through its API. The reasoning model supports tool calls and a 128K-token context window, with a company-reported median time to first answer token below 320 milliseconds.
Inception