Skip to main content

TensorCost

The control plane for every dollar your company spends on AI. TensorCost is AI cost governance across self-hosted GPU fleets, managed-inference APIs (Bedrock, Azure OpenAI, Vertex, OpenAI, Anthropic), and the agent workloads that drive both. One pane of glass. Real attribution. Recommendations that pay for the platform.

Start here

Get started

Sign up, connect your first cloud account, see your first recommendation in 48 hours.

Install agents

A Docker container with console-issued credentials. IMDSv2 auto-detect, opt-in HMAC-authenticated streaming.

Connect Bedrock

The lead managed-inference adapter. Routing, prompt cache, provisioned-throughput, runaway-loop alerts.

Connect cloud accounts

AWS, Azure, and GCP infrastructure spend — separate from AI usage, read-only, verified before it activates.

Coverage matrix

TensorCost is the only governance layer that spans all three sides of the modern AI bill.

Three SKUs

GPU FinOps

The original product. Fleet rightsizing, MIG slicing, spot blending, RI/Savings Plan optimization.

AI Inference Cost Governance

Bedrock + Azure OpenAI + Vertex + OpenAI + Anthropic, with model routing and prompt-cache recommenders.

Agent Cost Observability

Per-agent, per-workflow attribution. Loop and retry-storm anomaly detection. PagerDuty + Slack hooks.

Explore

Architecture

15 backend services, 14 microfrontends, 21 packages. NestJS + gRPC + Postgres + Redis on Fargate.

API reference

REST gateway, JWT-auth, tenant-scoped, versioned at /v1/.

CLI

@tensorcost/cli — offline prompt-cost estimates and before/after comparisons from your terminal or CI.

Developer tools

@tensorcost/sdk (Node) and tensorcost (Python) — one-line wrappers for OpenAI, Anthropic, and Bedrock clients.

Real-time events

Tenant-scoped socket.io for live recommendations, alerts, and agent fleet state.

ML & anomaly detection

Burn-rate alerts, forecasts, and the four shipped Bedrock recommenders.

Observability

OpenTelemetry, structured logs, customer-side audit trails.

Configuration

RBAC, tag mapping, budget hierarchies, alert routes, feature flags.

Feature flags

LaunchDarkly + useFeature(). Quarterly stale-flag cleanup ritual.

Compliance & SOC 2

Not yet SOC 2 certified — here’s what’s in place today and what’s on the roadmap. RLS on 21+ tables.

Built by Vaadhlabs

TensorCost is built and operated by Vaadhlabs. Reach the team at support@tensorcost.com — our source repository is private, so email is the right way to reach us for now.