Nexus Composer
An AI coworking agent that works alongside you on every task
Development roadmap 2026–2027
Shaping the next step with OneNexus
Inference and Nexus Composer are open for pilot sign-ups, and Training is in private preview. Expanding GPU capacity and building tools for agents are our next goals.
Development roadmap
Product journey
-
Now
Inference
Two ways to run inference, one unified API standard
Training
Train and deploy models, end to end
-
Dec 2026
Semantic Routing
Automatically route each request to the lowest-cost model that still keeps response quality
Who it’s for Mixed workloads
-
May 2027
GPU Serverless
Run AI workloads flexibly against real traffic, with no fixed GPU cluster to maintain
Who it’s for Event-driven workloads
AI Sandbox for agents
An isolated environment to test agent tools, workflows and behaviour safely
Who it’s for Agent development teams
GPU Rental
Reserve dedicated GPU capacity ahead of time for workloads that run continuously
Who it’s for Steady resource needs
-
Date to be announced
Neo Cloud
Extended AI cloud infrastructure that brings GPU Serverless and GPU Rental together
Who it’s for Organisations that need their own AI infrastructure
Data Platform
Standardise how data is collected, integrated, governed and put to work for AI, in one place
Who it’s for Data & AI teams
Roadmap dates reflect the current plan and may change. Subscribers receive changes.
Stay informed
Follow the OneNexus development roadmap
Hear first about product release milestones, early-access opportunities and how our infrastructure is expanding.