Get your limits
Returns the org's quotas, with what is in use against each, and the
caller's explicitly configured limits, which only an agent
credential has in this version. An org's quotas cap its live
on-demand instances and their vCPUs, stopped ones included, and not
GPU instances, which are reserved or spot, and its instances with a
dedicated IPv4, GPU instances included; one with no quota,
such as the platform's own, has inUse and no limit. Creates
still need available host resources. org.clusters is present only for an org with managed
Kubernetes. It also says whether the caller
counts as an agent, whose GPU reservations wait for an owner's
approval. Any valid credential can
read it. Plan a create from it, rather than retrying one that
returned 403 quota_exceeded.
AuthorizationBearer <cc_...>An API key of your org, sent as Authorization: Bearer cc_.... Keys
start with cc_ and are otherwise opaque. Treat a key as a password:
anyone holding it can act with its scopes until it expires or is
revoked. Each operation's security requirement lists the scopes it
needs; an empty list means any valid key. The MCP server at /mcp
accepts agent keys and no other kind.
Api-Version*stringThe API version the request is written against. This document
describes 2026-10-15. A missing header returns
400 api_version_required, and a version the API does not serve
returns 400 unsupported_api_version.
X-Correlation-ID?stringAn id of your own for this request, or for a group of requests, to
find them in our logs and yours. Logged with the request and returned
unchanged in X-Correlation-ID. The request id itself is always the
API's own, in X-Request-ID.
^[A-Za-z0-9._:-]+$1 <= length <= 128Org-Context?stringAdmin keys only: the id of the org to act in. Required with an admin
key, together with Access-Reason; with any other credential it
returns 400 invalid_request. Every request that carries it is
recorded in that org's operator-access log.
^org-[a-z0-9]([-a-z0-9]*[a-z0-9])?$5 <= length <= 63Access-Reason?stringAdmin keys only: why the operator is acting in the org, such as a
support ticket. Required with Org-Context. Recorded, verbatim, in
the org's operator-access log.
^[^\x00-\x1F\x7F]+$10 <= length <= 500Your limits, and what is in use against each.
application/json- response
What a create is checked against, and what is in use against it. A limit that is absent does not apply.
agent*booleanWhether the caller counts as an agent: an agent key, or a token
issued to any client but our console. An agent's instances
always expire, and its GPU reservations wait for an owner's
approval.
org*The org's quotas, which hold every caller in it, and what is in use against each: its live on-demand instances and their vCPUs, from create to delete, stopped ones included. GPU instances, reserved or spot, do not count toward those two. Dedicated IPv4s count every instance. A new org has a small quota, raised once it has paid an invoice.
caller*The caller's explicitly configured limits: an
agent key's, set when the key was made, or, for a token issued to
an agent, the org's defaults for agents. What counts against them is
what the caller created that is still live: with that key, or with
tokens the same person got through the same client. A limit that is
absent does not apply; a person signed in to our console, and a
standard key, have none.
curl -X GET 'https://api.cloud.confidential.ai/v1/limits' \ -H 'Api-Version: 2026-10-15' \ -H 'X-Correlation-ID: rollout-2026-10-22' \ -H 'Access-Reason: Support ticket 4821: customer asked why i-6d3f9h2k8mzr failed'{ "agent": true, "org": { "id": "org-acme", "instances": { "limit": 20, "inUse": 18 }, "vcpus": { "limit": 64, "inUse": 52 }, "dedicatedIpv4": { "limit": 5, "inUse": 1 } }, "caller": { "instances": { "limit": 8, "inUse": 3 }, "vcpus": { "limit": 16, "inUse": 12 }, "sizes": { "allowed": [ "c1.2x", "c1.4x", "s1.2x", "s1.4x" ] }, "locations": { "allowed": [ "us-west-1" ] } }}List notifications GET
Lists what your org was sent in the last 30 days, newest first, with each channel's delivery: pending while it is retried, `delivered` once the receiver answered 2xx, `failed` after a day of retries. Requires `alerts:read`.
Get your usage GET
Returns what your org used in each period from `start` until `end`, a UTC day unless `granularity` says otherwise, by meter and location, and by inference model and key when `groupBy` names them, from the hours closed so far. Compute counts while an instance's status is `running`: vCPUs and memory for a CPU size, GPUs for a spot GPU size, and nothing for a size whose reservation is paid for. Disks count from the instance's first `running` until it fails or its delete is asked for, so a create that never runs costs nothing, a dedicated IPv4 address while it is held, and egress as the bytes the instance sends. An hour closes 15 minutes after it ends and never changes after, so the totals are final up to `closedThrough`. Inference counts each request's tokens in the hour it finished: uncached input, cached input and output, each a meter of its own. A period, meter and location with no use is left out. Any valid credential can read it.