Skip to main content
POST
Python (SDK)

Autorizaciones

Authorization
string
header
requerido

Gateway auth: send Authorization: Bearer <mka1-api-key>. For multi-user server-side integrations, you can also send X-On-Behalf-Of: <external-user-id>.

Encabezados

X-On-Behalf-Of
string

Optional external end-user identifier forwarded by the API gateway.

Parámetros de ruta

run_id
string
requerido

Cuerpo

application/json
trials
object[]

Cada ensayo fija su propio sample_index (único por tarea y modelo): volver a enviar un lote reemplaza las mismas filas en lugar de duplicarlas, por lo que un lote reintentado es seguro.

Maximum array length: 1000
finalize
boolean
predeterminado:false

Marca la ejecución como completada una vez que estas pruebas se ejecutan — la última llamada de un trabajo de streaming envía esto, normalmente con un trials vacío.

Respuesta

200 - application/json

Está bien

id
string
requerido
object
any
requerido
suite_id
string
requerido
suite_version
integer
requerido
Rango requerido: -9007199254740991 <= x <= 9007199254740991
suite_version_id
string
requerido
org_id
string
requerido

The org that owns this run.

team_id
string
requerido

The team that owns this run.

status
enum<string>
requerido
Opciones disponibles:
queued,
in_progress,
finalizing,
completed,
failed,
cancelling,
cancelled
models
string[]
requerido
task_ids
string[] | null
requerido
judge_model
string | null
requerido
embedding_model
string | null
requerido
generation
object
requerido
request_counts
object
requerido
metrics
object | null
requerido
error
object | null
requerido
artifact_file_ids
string[]
requerido
metadata
object
requerido
agent_name
string | null
requerido

Coding-agent harness that produced this run (e.g. omp, claude-code). Null for every non-harness eval kind — a null here is not a missing value, it means the run is not an agent run.

agent_version
string | null
requerido

Version of agent_name, as the harness reported it. Null when agent_name is.

agent_effort
string | null
requerido

Reasoning-effort setting the agent ran at (e.g. medium, xhigh). Part of the leaderboard's row identity: the same agent and model at two efforts are two rows, not one. Null when agent_name is.

cost_usd
number | null
requerido

Total spend for the run in USD. Null when the harness did not report cost.

tokens
object
requerido

Token totals for the run, summed over its samples when it finalized. Each is null when nothing reported that figure, which is a different fact from zero and must not render as free.

created_at
integer
requerido
Rango requerido: -9007199254740991 <= x <= 9007199254740991
started_at
integer | null
requerido
Rango requerido: -9007199254740991 <= x <= 9007199254740991
completed_at
integer | null
requerido
Rango requerido: -9007199254740991 <= x <= 9007199254740991
cancelled_at
integer | null
requerido
Rango requerido: -9007199254740991 <= x <= 9007199254740991
failed_at
integer | null
requerido
Rango requerido: -9007199254740991 <= x <= 9007199254740991