Skip to main content
POST
/
calls
/
stream_query
[
  {
    "id": "<string>",
    "project_id": "<string>",
    "op_name": "<string>",
    "display_name": "<string>",
    "trace_id": "<string>",
    "parent_id": "<string>",
    "thread_id": "<string>",
    "turn_id": "<string>",
    "started_at": "2023-11-07T05:31:56Z",
    "attributes": {},
    "inputs": {},
    "ended_at": "2023-11-07T05:31:56Z",
    "exception": "<string>",
    "output": "<any>",
    "summary": {},
    "wb_user_id": "<string>",
    "wb_run_id": "<string>",
    "wb_run_step": 123,
    "wb_run_step_end": 123,
    "deleted_at": "2023-11-07T05:31:56Z",
    "storage_size_bytes": 123,
    "total_storage_size_bytes": 123
  }
]

Authorizations

​
Authorization
string
header
required

Basic authentication header of the form Basic <encoded-value>, where <encoded-value> is the base64-encoded string username:password.

Headers

​
accept
string
default:application/jsonl

Body

application/json
​
project_id
string
required
​
filter
object | null
​
limit
integer | null
​
offset
integer | null
​
sort_by
SortBy · object[] | null
​
query
object | null
​
include_costs
boolean | null
default:false

Beta, subject to change. If true, the response will include any model costs for each call.

​
include_feedback
boolean | null
default:false

Beta, subject to change. If true, the response will include feedback for each call.

​
include_storage_size
boolean | null
default:false

Beta, subject to change. If true, the response will include the storage size for a call.

​
include_total_storage_size
boolean | null
default:false

Beta, subject to change. If true, the response will include the total storage size for a trace.

​
columns
string[] | null
​
expand_columns
string[] | null

Columns to expand, i.e. refs to other objects

Examples:
[
  "inputs.self.message",
  "inputs.model.prompt"
]
​
return_expanded_column_values
boolean | null
default:true

If true, the response will include raw values for expanded columns. If false, the response expand_columns will only be used for filtering and ordering. This is useful for clients that want to resolve refs themselves, e.g. for performance reasons.

Response

Stream of data in JSONL format

​
id
string
required
​
project_id
string
required
​
op_name
string
required
​
trace_id
string
required
​
started_at
string<date-time>
required
​
attributes
object
required
​
inputs
object
required
​
display_name
string | null
​
parent_id
string | null
​
thread_id
string | null
​
turn_id
string | null
​
ended_at
string<date-time> | null
​
exception
string | null
​
output
any
​
summary
object
​
wb_user_id
string | null
​
wb_run_id
string | null
​
wb_run_step
integer | null
​
wb_run_step_end
integer | null
​
deleted_at
string<date-time> | null
​
storage_size_bytes
integer | null
​
total_storage_size_bytes
integer | null