Sitelet https://docs.firecrawl.dev/api-reference/endpoint/agent-get
Skip to main content
GET
Get the status of an agent job
Are you an AI agent that needs a Firecrawl API key? See firecrawl.dev/agent-onboarding/SKILL.md for automated onboarding instructions.

Authorizations

Authorization
string
header
required

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

Path Parameters

jobId
string<uuid>
required

The ID of the agent job

Response

Successful response

success
boolean
status
enum<string>
Available options:
processing,
completed,
failed
data
object

The extracted data (only present when status is completed). Never set on a failed run, including one that stopped at its credit limit.

partial
any

Best-effort JSON recovered from an incomplete run. Only present on some failed runs, such as one that reached maxCredits; never present while the run is processing. It may be missing rows or fields, so do not treat it as a completed result.

partialSchemaValid
boolean

Whether partial validates against the request's JSON schema. Only present when partial is present and the request included a schema.

stopReason
enum<string>

Why a failed run stopped early. credit_limit_reached means the run hit its maxCredits limit. Only present when status is failed.

Available options:
credit_limit_reached
message
string

A note from the agent about the result. In chat mode, a follow-up that asks for no new data is answered here instead of in data. On a failed credit-limit stop, it may explain what is incomplete.

model
enum<string>
default:spark-2

Model preset used for the agent run. Every new run executes on spark-2; Spark 1 names only appear on legacy runs.

Available options:
spark-2,
spark-1-pro,
spark-1-mini
effort
enum<string>

Reasoning budget used for the agent run (only present for runs that set effort)

Available options:
low,
medium,
high
error
string

Error message (only present when status is failed)

expiresAt
string<date-time>
creditsUsed
number
threadId
string<uuid>

The thread this run belongs to.

threadTurn
integer

This run's turn in the thread, starting at 1.

mode
enum<string>
Available options:
extract,
chat
suggestions
object[]

Follow-ups the agent offers for the next turn of the thread.

pendingApproval
object

What the turn stopped to ask. Answer it on the next turn of the thread with exchange.approve or exchange.decline.

exchange
object

What the run did with Alexandria. toolkits and requireApproval are what the run resolved to after thread inheritance.