Skip to main content
GET
Get Available Fft Models

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Query Parameters

team_id
string | null

Optional team ID. When set, returns FFT models cached on the team's eligible PrimeClusters. When omitted, uses the caller's personal allocations.

Response

Successful Response

What the caller can dispatch to, across both capacity paths.

models is reserved capacity: repos pre-cached on clusters the caller holds a ClusterAllocation on, sorted by model name. Same principal collapse as the dispatch picker (team wins when present, otherwise personal). Only PRESENT-cache clusters contribute — ABSENT / ERROR clusters would 400 the dispatch, and NULL (unprobed) clusters haven't reported a manifest yet.

onDemand is the shared pool, which needs no allocation and bills per GPU-hour. It is populated regardless of whether the caller has any reserved capacity, because a caller with none still needs to see what they can buy — an empty response was previously the only signal, and it read as "nothing is running".

models
AvailableFFTModel · object[]
onDemand
OnDemandGpuTypeAvailability · object[]
onDemandBetaAccess
boolean
default:false

Whether this account is enrolled in the on-demand beta. False means onDemand is empty because of enrolment, not because capacity is missing — it lets a client say 'ask for access' instead of 'nothing is running'. Carries no rates, so a non-enrolled account still learns nothing about pricing.