IQuest released IQuest-Q1, a sparse Mixture-of-Experts agentic foundation model with 320B total parameters and 15B activated parameters, built for agentic CLI systems. The model is trained in three stages, beginning with pre-training that combines Stable SplitNorm, Muon and muP under a two-stage curriculum. The release details the architecture and training approach for developers building command-line AI agents.
We present IQuest-Q1, a sparse Mixture-of-Experts (MoE) agentic foundation model with 320B total parameters (15B activated). IQuest-Q1 is developed through a three-stage process: (1) stable and efficient pre-training, combining Stable SplitNorm, Muon, and muP with a two-stage curriculum whose second
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.