TrueFoundry
7.2/10TrueFoundry is a Kubernetes-based AI deployment platform and AI gateway that lets enterprise teams deploy models and govern LLM, MCP and agent traffic in their own cloud.
Data last updated November 2026 · Our methodology
Overview
TrueFoundry started as an MLOps platform that hides Kubernetes behind a developer-friendly layer. Teams connect a cluster in their own cloud or on-premises to a hosted control plane, then deploy models as autoscaling APIs, run training and batch jobs on a schedule, open cloud notebooks, and fine-tune or serve open-source LLMs. A split-plane design keeps workloads and data in the customer's environment, so services keep running even if TrueFoundry's control plane does not. The company now leads with its AI Gateway, which routes requests to LLM providers and MCP servers and adds fallbacks, rate limits, budgets, guardrails and observability. It suits mid-size and enterprise data science and platform teams that want to self-host AI on AWS, Azure or GCP without hiring a large DevOps team. Pricing has three tiers. Developer is free for up to 3 users. Pro costs $25 per user per month and includes 20K gateway requests per user, with overage at $20 per additional 100K requests. Enterprise has custom pricing and adds VPC, on-prem, air-gapped and data-residency deployment along with SCIM and audit logs. Model provider token charges are billed separately, and TrueFoundry says it does not mark them up. It integrates with Kubernetes, Terraform, MLflow, Hugging Face, Datadog, Grafana and Prometheus. The main limitation is that TrueFoundry does not supply compute or hosting itself. You run and pay for the underlying cloud infrastructure, and teams without Kubernetes experience report a learning curve.
Our Rating
AI Software7.1/10
Hosting & Infrastructure7.3/10
Who TrueFoundry is best for
Best software for EnterpriseGood fit
It offers VPC, on-prem and air-gapped deployment along with SSO, SCIM, audit logs, and SOC 2 and HIPAA compliance.
ML engineers can deploy models and jobs without learning Kubernetes, though the platform has a learning curve.
Reviewers say it suits mid-size data science teams that lack dedicated DevOps staff.
Less suited
There is a free tier for 3 users, but reviewers note that cost is a concern for smaller teams.
Pros & Cons
Pros
- The split-plane architecture keeps workloads and data inside the customer's own cloud or on-prem cluster.
- Data scientists can deploy autoscaling model APIs and scheduled jobs without writing Kubernetes manifests.
- G2 reviewers rate it about 4.6 out of 5 and often praise how quickly the support team responds.
- Compliance coverage includes SOC 2, HIPAA and GDPR, with air-gapped options on Enterprise.
- The free Developer tier and a $25 per user Pro plan make the gateway easy to trial.
Cons
- It provides no compute of its own, so customers pay their cloud provider separately for clusters and GPUs.
- Teams without Kubernetes skills report a learning curve and need more onboarding.
- Reviewers mention documentation gaps and parts of the product that feel early-stage.
- Enterprise pricing, and pricing for the deployment platform and Agent Gateway, is not public.
Key Features
TrueFoundry Pricing
Developer
- 7-day log retention
- Up to 3 users and 2 teams
- 5 model accounts and 5 MCP servers
- 10K AI Gateway requests per user per month
- Support through Discord and an in-app chatbot
Pro
or $300/user/year
- Support portal
- 30-day log retention
- SSO and custom roles
- Up to 10 budget limit rules
- Unlimited users and up to 10 teams
- 20K gateway requests per user per month
- Built-in guardrails, with up to 10 policies
Enterprise
- Custom log retention
- SCIM, audit logs and custom guardrails
- Priority support with an SLA and a dedicated Slack channel
- SaaS, VPC, on-prem, air-gapped and data-residency deployment
Frequently Asked Questions
Yes. The Developer plan is free for up to 3 users and includes 10K gateway requests per user each month.
No. Workloads run on a Kubernetes cluster in your own cloud account or on-premises, connected to TrueFoundry's control plane.
No. Model provider charges are billed separately by the provider, and TrueFoundry says it does not mark up tokens.
The vendor cites SOC 2, HIPAA and GDPR, and RFP.wiki also lists SOC 2 Type 2 and ITAR.
Extra gateway requests cost $20 per additional 100K, pooled across the team.
Yes, on the Enterprise plan. Enterprise also supports VPC, on-prem and air-gapped deployments.
Similar Tools
See all alternatives
Amazon's public cloud platform, with over 200 pay-as-you-go services for compute, storage, databases, networking and AI, used by startups through large enterprises and governments.
Portkey is an AI gateway and LLMOps control plane that routes, monitors, and governs requests to 1,600+ LLMs for engineering teams and enterprises.
Open-source and enterprise system for running, scheduling and scaling AI training, inference and batch jobs across Kubernetes, Slurm, on-prem and 20+ clouds.
Ranked near TrueFoundry in AI Software
See TrueFoundry in the listTrueFoundry is #59 in AI Software by our score.
