toolcompass.

Find your next AI tool

Search by product name or task. Press Escape to close.

AUDR

● Identity checked

AI usage infrastructure · Official vendor documentation reviewed 2026-10-07

AUDR is an open usage-record standard for attributing AI-agent infrastructure costs. It gives providers and developers a shared JSON format for identity, run, span and resource usage, with SDKs and adapter examples for capturing records and passing them to billing systems. It measures usage rather than providing a generative model.

Updated 2026-10-07View sources
Official website
Category
AI usage infrastructure
Free access
Apache-2.0 open standard and SDK code; infrastructure and billing services are separate.
API access
Python audr and TypeScript @openaudr/audr SDKs; adapters and billing sinks documented.

Is AUDR right for you?

A good fit for

Teams connecting multi-step agent runs to metered usage and billing records.

Before you choose

Requires instrumentation and a compatible sink. A shared run identifier helps attribution but does not establish final cost accuracy; validate schemas, aggregation and billing rates. No hosted-model entitlement is included.

Instrument a bounded agent run

Assign a run identifier, emit a record for each relevant operation and validate it against the schema. Compare joined usage with provider records before feeding a billing sink. Treat conformance fixtures as format validation rather than proof that an invoice is correct.

Open standard and billing handoff

AUDR’s specification and SDKs are Apache-2.0. Model tokens, gateway hosting, storage and billing platforms remain separate costs. Chargebee stewards the project; the PH launch title refers to this usage standard rather than a new generative Chargebee product.

What it can do

Features & capabilities

Unknown is different from unavailable. Each fact carries its own evidence.

CapabilityValueEvidenceChecked
StandardVersion 1.0.0 usage record and JSON Schema Draft 2020-12.Facts sourced2026-10-07
AttributionRun/span identity joins records from multiple metered operations.Facts sourced2026-10-07
AdaptersExamples cover LiteLLM, Vercel AI, Mastra and gateway integrations.Facts sourced2026-10-07
Data boundaryRecords describe usage; the specification excludes prompts, key material and PII.Facts sourced2026-10-07

Understand the total cost

AUDR pricing & plans

Pricing

Not listed

Official monthly price not confirmed

API
Python audr and TypeScript @openaudr/audr SDKs; adapters and billing sinks documented.
Explore pricing & history

Alternatives to AUDR

View all ↗

The practical questions

Frequently asked questions

Does AUDR run an AI model?

No. It standardizes metered usage records from agent operations.

Do records contain prompts?

The official standard says prompt text, key material and PII are excluded from the usage record.

What is AUDR used for?

AUDR is an open usage-record standard for attributing AI-agent infrastructure costs. It gives providers and developers a shared JSON format for identity, run, span and resource usage, with SDKs and adapter examples for capturing records and passing them to billing systems. It measures usage rather than providing a generative model.

Does AUDR have a free plan?

Apache-2.0 open standard and SDK code; infrastructure and billing services are separate.. This record does not confirm an ongoing free plan.

How much does AUDR cost?

A listed monthly price has not been confirmed. See the plan cards for entitlements, billing commitments and seat minimums.

Can I use AUDR through an API?

Python audr and TypeScript @openaudr/audr SDKs; adapters and billing sinks documented.. API access and subscription access may have different terms; consult the linked sources.

What should I check before choosing it?

Requires instrumentation and a compatible sink. A shared run identifier helps attribution but does not establish final cost accuracy; validate schemas, aggregation and billing rates. No hosted-model entitlement is included.

Price history

No retained pricing changes yet. A current price alone does not establish a historical trend.

How this profile is supported

Official product documentation

Vendor website reviewed for current workflow and limitations; no application test performed.

Read original source ↗

Facts apply to the named version and check date. Send a sourced correction if something changed.