ioMoVo has released V2 of its enterprise AI platform, giving teams tools to build agents, choose models and keep their content in its existing storage. The headline change is control: organisations can route work between models and run the platform in air-gapped environments.
Watch Desk analysis
What happened
The release adds ioStudio, a workspace for building agents, tools and playbooks, with a visual prompt builder. Teams can set an ordered list of models for different tasks, with requests passed to the next model if one fails or reaches its limit. ioMarket connects to 11 model providers, including Claude, Gemini, ChatGPT, Cohere and Groq.
Agents can work with content in services such as Box, Google Drive, SharePoint and S3-compatible storage without requiring migration, according to ioMoVo. The release also describes support for MCP and Agent2Agent, plus air-gapped deployments with what the company calls full feature parity. Read the ioMoVo V2 announcement.
Our top picks
- Model routing
Teams can set an ordered fallback list for each type of work, rather than relying on one model to behave forever. - Bring agents to existing content
The platform connects to several storage systems without requiring organisations to move their files first. - Air-gapped deployment
ioMoVo says the platform can run without an internet connection, with full feature parity. - Connect other agents
MCP and Agent2Agent support lets other agents draw on the same governed content. - Build and tailor agents
ioStudio includes a visual Prompt Builder and agents for indexing images, video, documents and audio.
Why it matters
The practical question for organisations adopting AI is not just which model answers best. It is also where data goes, what agents can access, and what happens when a model is unavailable. ioMoVo’s pitch is to put those decisions in the organisation’s hands, including the option to keep deployment air-gapped.
That could matter to customers with strict data-location or connectivity requirements. The release describes product capabilities, not independent evidence of how they perform in customer deployments, so the useful test will be what teams can actually configure and operate.
Our read
This is a substantial enterprise-platform update, not just another agent demo: the combination of model routing, existing-content connections and air-gapped deployment is the story. The promise is appealing, though “your agents, your choice” only goes so far if the controls prove fiddly in practice. Buyers should ask for a hands-on demonstration of routing, permissions and failure handling before letting an agent near the filing cabinet.
What to watch
- Whether customers can configure model routing and fallback behaviour without specialist help.
- How permissions and governance work when agents connect to existing content.
- Which deployment options and integrations are available in practice.
Discussion spark: For organisations handling sensitive information, is the ability to choose models and keep content in-house enough to make agent platforms trustworthy, or should independent deployment evidence be a prerequisite?
Sources and evidence
- ioMoVo V2 Lets Teams Build Their Own AI Agents and Choose the Models Behind Them PR Newswire – Chinook Observer (17 November 2026, 18:00 UTC)
- Isomorphic Labs announces $5 Billion Series C funding – PR Newswire (17 November 2026, 18:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.