# TrueFoundry > Enterprise AI gateway for routing, governing and observing LLM, MCP and agent traffic in your own infrastructure. TrueFoundry is an enterprise AI gateway that sits between applications and the language models, MCP servers and agents they use. TrueFoundry provides one interface for many model providers and adds routing, rate limits, guardrails, cost controls and tracing. TrueFoundry is used as a hosted service or installed in a company's own cloud or data centre. **Best for:** Engineering and platform teams that need one governed, observable access layer for LLMs, MCP tools and agents, including in private or regulated environments. - Official site: https://www.truefoundry.com - Listed in: [DevOps & Monitoring](https://trendhub.net/ai-tools/category/devops-monitoring), [AI Agents & Automation](https://trendhub.net/ai-tools/category/ai-agents), [Developer API](https://trendhub.net/ai-tools/category/developer-api) - Last checked: 2026-10-09 (facts taken from truefoundry.com) - Source page: https://trendhub.net/ai-tools/truefoundry-enterprise-ai-gateway-llms-mcp ## Key facts - Platforms: Web, API - Pricing model: Freemium - Free plan: Yes - Free trial: Yes - API: Yes - Made by: Ensemble Labs Inc ## What is TrueFoundry? TrueFoundry is an enterprise platform that places a single gateway between applications and the AI models, tools and agents they call. Its LLM Gateway offers one interface to many model providers with routing, fallbacks, rate limits and cost controls. An MCP Gateway handles discovery, authentication and access rules for MCP servers and tools, and an Agent Gateway applies identity, usage policies and tracing to agent activity. Around these sit prompt management, a skills and agent registry, guardrails, budget limits and audit logs. Observability is OpenTelemetry-compliant and can feed Grafana, Datadog or Prometheus. The platform is sold as SaaS and can also be installed in a VPC, on-premises, in air-gapped setups or across several clouds. Other products cover model deployment, training and fine-tuning. The company behind it is Ensemble Labs Inc of San Francisco. ## Key features - **LLM Gateway**: Sends requests to different model providers through one endpoint and applies routing rules, fallbacks, rate limits and spending caps. - **MCP Gateway**: Registers MCP servers centrally and controls which agents can reach which tools, with authentication and an audit trail. - **Agent Gateway**: Gives each agent a governed identity and traces what it calls, under policies set by administrators. - **Guardrails and budgets**: Enforces usage quotas, data residency rules and cost limits while workloads are running. - **Tracing and monitoring**: Records latency, throughput, token usage, cost and GPU utilisation and exports them to existing observability tools. - **Access control and audit logs**: Supports single sign-on, role-based permissions and a log of model usage, user access and configuration changes. ## Use cases - **Centralising model access**: A platform team gives every internal application one governed route to several LLM providers. - **Controlling AI spend**: Finance and engineering set budget rules per team and watch token costs on shared dashboards. - **Securing agent tool use**: Security staff decide which MCP tools agents may call and require approvals for sensitive ones. - **Running AI inside a private network**: A regulated organisation installs the platform in its own VPC or on-premises so data stays in its domain. - **Auditing AI activity**: Compliance officers export logs of who used which model and what was changed. ## Things to know - **Seats and traffic are billed separately**: Plans include a pooled monthly allowance of requests and tool calls per user; traffic beyond it is charged as overage rather than by adding users. - **What counts as a request**: Every call to a model and every MCP tool invocation is counted once, while retries and failed calls inside the gateway are not counted. - **Model costs are extra**: Fees from model providers and managed third-party guardrails are not part of the subscription. - **Some products are quote-only**: The pricing FAQ says the Agent Gateway and Agent Harness are custom priced for now. ## FAQs ### What is TrueFoundry? TrueFoundry is an enterprise AI gateway and deployment platform. It routes and governs traffic to LLMs, MCP servers and agents and records it for monitoring. ### What is TrueFoundry used for? TrueFoundry is used to give teams a single controlled entry point to AI models and tools, with routing, rate limits, guardrails, cost tracking and audit logs. ### Is TrueFoundry free? TrueFoundry has a Developer plan described as free forever for up to three users, with a monthly request allowance per user. Larger teams move to the Pro or Enterprise plans. ### How much does TrueFoundry cost? TrueFoundry prices its Pro plan per user per month with overage charged on extra gateway requests, and Enterprise is custom. See the official pricing page for current figures. ### Can TrueFoundry be self-hosted? Yes. TrueFoundry can run in a VPC, on-premises, in air-gapped environments or across multiple clouds, in addition to its SaaS option. The plan comparison places VPC and on-premises deployment in the Enterprise tier. ### Are model provider costs included in TrueFoundry? No. TrueFoundry's pricing FAQ says the fees of model providers (Anthropic, OpenAI, AWS, Azure, Google and others) are billed separately. ### Who makes TrueFoundry? TrueFoundry's site names Ensemble Labs Inc, based in San Francisco, as the company behind the product. --- Source: https://trendhub.net/ai-tools/truefoundry-enterprise-ai-gateway-llms-mcp (trendhub)