Agent report card
Can an agent use OpenRouter API through Model access lifecycle?
Grade B+: The public commitment omits a material component until later in the flow.
OpenRouter / active
Five-stage assessment
Where agent autonomy holds and where it breaks
- EvaluateReady
- Sign upReady
- PayLimited
- ProvisionReady
- OperateReady
The exact service and stable evaluation or access entrypoints are publicly discoverable.
Applicable terms are stable, readable, retrievable, and materially complete.
Eligibility conditions and required inputs are explicit and decidable.
Price or no-charge status, variables, currency, and material conditions are explicit.
Verified structured artifacts accelerate discovery of the evaluation surface.
Web-agent access policy is explicit and verified for the assessed resources.
Evidence and actions6 signals
Can an agent decide every material eligibility condition before commitment?
supported: freshEligibility conditions and required inputs are explicit and decidable.
OpenRouter specifies eligibility requirements in Section 2 of its Terms of Service, requiring users to be at least 13 years of age (with parent/guardian permission if under 18) and requiring representative authority to bind an organization.
Can an agent estimate cost or confirm no-charge status before commitment?
supported: freshPrice or no-charge status, variables, currency, and material conditions are explicit.
OpenRouter explicitly declares its pricing variables, currency (USD), rates, and conditions. Per-token prompt and completion rates are provided via the models API and browser, base currency is USD, and unused credits may expire after one year.
Can an agent find the exact service and its stable entrypoints?
supported: freshThe exact service and stable evaluation or access entrypoints are publicly discoverable.
OpenRouter provides public discovery artifacts, including llms.txt and documentation guides, that disclose stable API endpoints such as https://openrouter.ai/api/v1/chat/completions for service access.
Are verified structured discovery artifacts available?
supported: freshVerified structured artifacts accelerate discovery of the evaluation surface.
OpenRouter exposes structured evaluation and discovery artifacts, including an OpenAPI specification at https://openrouter.ai/openapi.json and an index at https://openrouter.ai/llms.txt.
Can an agent retrieve and understand the applicable commitment terms?
supported: freshApplicable terms are stable, readable, retrievable, and materially complete.
The applicable Terms of Service are retrievable, readable, and materially complete on https://openrouter.ai/terms, defining service use, model terms, authorized user obligations, and account requirements.
Does the vendor deliberately describe access for web agents?
supported: freshWeb-agent access policy is explicit and verified for the assessed resources.
OpenRouter explicitly declares crawler and web agent policy in robots.txt, allowing access to all paths under '/' while disallowing '/seo/'.
A stable canonical route begins the required access bootstrap.
Scoped revocable credentials or delegated authorization have explicit consent and resumption boundaries.
Evidence and actions5 signals
Can an agent operate the access controls and safe handoffs deterministically?
supported: freshCurrent admissible evidence does not resolve this finding.
The sign-up page captured states 'We could not find a primary email address on your account. Please contact support.' It does not contain evidence demonstrating that an agent can deterministically operate access controls, safe handoffs, validation, or navigation.
Is there a stable route to begin obtaining service access?
supported: freshA stable canonical route begins the required access bootstrap.
OpenRouter provides stable canonical entrypoints to obtain service access, including account sign-up ('Create your OpenRouter account to access hundreds of AI models') and an OAuth PKCE authentication guide for connecting users.
Can an agent obtain access without an unsupported CAPTCHA boundary?
supported: freshCurrent admissible evidence does not resolve this finding.
The retained initial sign-up document, crawler policy, and OAuth guide do not establish whether a mandatory CAPTCHA appears in later registration branches.
Can an agent obtain scoped, revocable authority for this service?
supported: freshScoped revocable credentials or delegated authorization have explicit consent and resumption boundaries.
OpenRouter provides explicit programmatic mechanisms to obtain scoped, revocable API key credentials via POST /api/v1/keys, including parameters for expiration, limit bounds, and endpoints to list, get, update, or delete keys.
Can required phone verification be completed through a supported boundary?
supported: freshCurrent admissible evidence does not resolve this finding.
The retained initial sign-up document and OAuth guide do not establish whether phone verification appears in later registration branches or how such a handoff resumes.
The public commitment omits a material component until later in the flow.
A documented direct purchase path reaches paid access without a vendor decision point.
Evidence and actions4 signals
Can an agent construct checkout, hand off approval safely, and resume?
supported: freshCurrent admissible evidence does not resolve this finding.
OpenRouter documentation describes adding credits via Stripe or crypto, but the provided evidence does not demonstrate an API or web flow supporting deterministic checkout construction, handoff, and resumption.
Is the exact commercial commitment disclosed before authorization?
supported: freshThe public commitment omits a material component until later in the flow.
OpenRouter discloses US dollars as the base currency and usage rates per million tokens on the models page, but notes that crypto payments are charged an unspecified fee amount ('Crypto payments are charged a fee of .').
Can payment be authorized within scoped agent or explicit human authority?
supported: freshCurrent admissible evidence does not resolve this finding.
While Stripe receipts and accepted payment methods (credit cards, AliPay, USDC) are mentioned, the retained evidence does not document a payment rail supporting scoped agent delegation or explicit human-confirmed authorization.
Is a direct self-service path to paid access documented?
supported: freshA documented direct purchase path reaches paid access without a vendor decision point.
To get started, users can create an account and manually add credits or set up auto top-up on the Credits page to access LLM inference without requiring vendor sales approval.
What would improve this stage
- Disclose the complete commitment before payment authorization.
Provisioning is machine-triggerable or follows deterministically from an allowed handoff.
Credentials, grants, or configuration are delivered through a documented agent-usable flow.
A synchronous success response returns a usable created resource or access material; otherwise asynchronous work exposes progress, terminal state, reconciliation, and a documented bound.
Evidence and actions3 signals
Can usable access material be delivered securely to an authorized agent?
supported: freshCredentials, grants, or configuration are delivered through a documented agent-usable flow.
OpenRouter API keys can be created via POST /api/v1/keys with a management key, returning the plaintext key string in the response data.
Can an agent determine successful provisioning completion and reconcile asynchronous failure?
supported: freshA synchronous success response returns a usable created resource or access material; otherwise asynchronous work exposes progress, terminal state, reconciliation, and a documented bound.
Creating a new API key synchronously returns a 201 response with the usable key string and created API key information object.
Can provisioning be initiated within supported agent authority?
supported: freshProvisioning is machine-triggerable or follows deterministically from an allowed handoff.
API key creation is machine-triggerable via a POST request to https://openrouter.ai/api/v1/keys using a Bearer authorization token.
Every essential target has a stable documented agent-usable interface alternative.
Request-time authentication is documented and usable under scoped agent authority.
Essential operations have stable readable request, response, and effect semantics.
Applicable limits, errors, retry, idempotency, cancellation, and reconciliation semantics are documented.
Credential lifecycle and recovery are documented and executable within delegated authority.
A verified agent-native interface covers the declared assessment targets.
Evidence and actions6 signals
Is an agent-native protocol interface verified against the essential targets?
supported: freshA verified agent-native interface covers the declared assessment targets.
OpenRouter explicitly provides an MCP server interface (mcp.openrouter.ai/mcp) that exposes tools such as list-models, get-model, get-credits, and send-message.
- inference-api
- model-catalog-api
- openrouter-mcp
Can an agent manage credential expiry, rotation, revocation, compromise, and recovery?
supported: freshCredential lifecycle and recovery are documented and executable within delegated authority.
OpenRouter provides API key lifecycle management endpoints to create keys with optional expiration dates and credit limits, list, retrieve, update, and delete keys.
Can an agent handle applicable failure modes safely?
supported: freshApplicable limits, errors, retry, idempotency, cancellation, and reconciliation semantics are documented.
OpenRouter documents standardized HTTP error status codes, mid-stream SSE error formats, typed error codes, retry policies using Retry-After headers, and availability fallback semantics.
Can an agent authenticate service operations with scoped authority?
supported: freshRequest-time authentication is documented and usable under scoped agent authority.
Operations are authenticated via HTTP Bearer tokens in the Authorization header using standard API keys or credentials obtained via OAuth PKCE flows.
Are essential operation inputs, outputs, and effects stable and decidable?
supported: freshEssential operations have stable readable request, response, and effect semantics.
All essential target endpoints (credits, error reference, models API, MCP server, OpenAPI spec, quickstart, rate limits, streaming, and tool calling) have documented schemas, parameters, and expected response payloads.
- inference-api
- model-catalog-api
- openrouter-mcp
Can an agent perform every essential assessment target through a usable interface?
supported: freshEvery essential target has a stable documented agent-usable interface alternative.
Every target resource and feature—including credits retrieval, error contract, live model listing, MCP server, OpenAPI specs, quickstart endpoints, rate limits, streaming, and tool calling—is accessible via documented HTTP API or MCP interfaces.
- inference-api
- model-catalog-api
- openrouter-mcp
What works
- Evaluate: The exact service and stable evaluation or access entrypoints are publicly discoverable.
- Sign up: A stable canonical route begins the required access bootstrap.
- Provision: Provisioning is machine-triggerable or follows deterministically from an allowed handoff.
- Operate: Every essential target has a stable documented agent-usable interface alternative.
What blocks agents
- Pay: The public commitment omits a material component until later in the flow.
Startup offers
OpenRouter on the startup-offers board
- Up to $5,000 in universal inference credits usable across all models on OpenRouter.OpenRouter for Startups
Improve the report
Correct it, rerun it, or improve the funnel.
- Profile
- arp_01m0j1a50vat933f76xwy3mec5
- Projection
- sha256:9af1f173ead60cd605a6c2b960b65fb5167d789f140c1c3a71fde98279b0e99e
- Policy
- service-use-2026-08-20-public-evidence-r9 · sha256:57bf95fd6d0fe6fe9173a68b60e8d8dc2179b357c88ff4349675a72d0e1d4139
Every request carries these exact identifiers.How report cards are assessed →View the exact declaration