Skip to main content
POST
Estimate a run

Authorizations

Authorization
string
header
required

Send Authorization: Bearer YOUR_API_KEY. The key resolves to a single organization.

Path Parameters

id
string<uuid>
required

Benchmark id.

Body

application/json
candidates
object[]
required

What the run compares, in display order.

Required array length: 1 - 10 elements
repeats
integer

Repeats per case (LLM, 1-3). STT always 1.

Required range: 1 <= x <= 3
versionNumber
integer

Published version to run. Default: the current published version (or the draft when nothing is published).

Required range: x >= 1
draft
boolean

Run the unpublished draft instead of a published version.

Response

Successful response.

A priced run plan plus the token that starts it.

estimateToken
string
required

Pass unchanged to start the run. Single use; expires in 15 minutes.

benchmarkId
string
required
versionNumber
number | null
required
isDraftRun
boolean
required
repeats
number
required
caseCount
number
required
cellCount
number
required
perCandidate
object[]
required
totalCostUsd
number
required

What Zelto bills (platform cost plus margin, plus bring-your-own-key fees).

warnings
string[]
required
planLimits
object[]
required

Plan limits starting this run would exceed. Non-empty adds the warning plan_limit; starting returns 403 plan_limit.

canStart
boolean
required

False when starting now would be refused (a plan limit, or spend blocked).

candidates
object[]
required
billing
object
required
spend
object
required
unverifiedReferences
integer

STT: cases scored against a reference transcript nobody has checked yet (warning unverified_references).

missingVariables
object | null

LLM: cases whose prompt references a {{variable}} they don't fill (warning missing_variables).

{key}
any