TrueFoundry, an enterprise AI infrastructure, today launched TrueForge, an open-source agent harness that gives enterprise teams a way to build, deploy, debug, and govern production AI agents without the vendor lock-in built into most managed agent platforms. TrueForge is an alternative to Claude Managed Agents and allows enterprises to build on any model or MCP server while reducing total agent operating costs by 50%.
Open models such as GLM-5.2 are rapidly closing the gap with frontier proprietary models at a fraction of the cost, yet most managed agent platforms still lock enterprises into a single vendor’s models, infrastructure, and pricing. The first wave of AI agents lived on individual developers’ laptops, inside coding tools and prototypes. The next wave is moving into customer-facing products and shared workflows, a shift that turns the agent harness underneath those products into a strategic control point. That’s because companies must manage persistent sessions, tool credentials, execution sandboxes, context, human approvals, debugging, access policies, and spending across every agent they operate. TrueFoundry is betting enterprises will want to own that layer rather than inherit it from a single model provider.
Founded in 2021 by IIT Kharagpur alumni Anuraag Gutgutia, Abhishek Choudhary and Nikunj Bajaj, TrueFoundry is backed by Peak XV’s Surge, Intel Capital, Eniac Ventures and Jump Capital, with teams based in India, San Francisco and Paris.
“I built AI infrastructure at Meta before starting TrueFoundry, and the pattern I kept seeing was teams trading control for convenience the moment they went from prototype to production,” said Nikunj Bajaj, co-founder and CEO of TrueFoundry. “Teams shouldn’t have to choose between a great agent-building experience and owning their AI stack. With TrueForge, you get that great experience, open source, on any model, with real enterprise governance, but at a fraction of the cost.”
Unlike platforms that only manage access at the workspace or resource level, TrueForge routes every model call and MCP interaction through TrueFoundry’s AI Gateway, so budget enforcement, rate limits, and guardrails can be applied to deliver a fully governed and secure managed agent experience for enterprises.
Teams can run TrueForge on their own infrastructure, bring their own models, MCP servers, and API keys, and route each task to whichever model fits the cost, latency, or quality needs of that job. TrueForge ships with support for OpenAI, Anthropic, and 20+ additional models, along with 40+ built-in tools, sandboxed execution, human-approval workflows, large-context handling, generative UI, and web search powered by Tavily. Paired with TrueFoundry’s AI Gateway and MCP Gateway, which already process more than 1 trillion tokens a day for enterprise customers, every model and tool call an agent makes is logged, governed, and policy-controlled. TrueFoundry is also launching a hosted, pay-per-usage version of TrueForge for teams that want the same experience without managing the infrastructure themselves.
Enterprises including Automatiq and NetApp are already running agentic workloads on TrueForge. In addition, TrueFoundry’s own Ask TFY, the conversational interface enterprise teams use to manage AI Gateway configurations and troubleshoot agent workflows, is built on the same harness.
“TrueFoundry has become the central platform in IT where agentic apps and agents are routed through. We are onboarding more users and teams every day, and it’s fundamentally changed how quickly we can get an agent from an idea to something running at scale,” said Robert Rubin, Senior Director of Platform Engineering at NetApp.
TrueForge is available on GitHub today at link. The open-source release is free to use with a team’s own model API keys. The hosted platform is self-serve, with usage-based pricing.