Skip to main content
POST
Run an inference query (streaming SSE)

Authorizations

Authorization
string
header
required

Static, non-expiring API key: Authorization: Bearer <FOURMINDS_API_KEY>.

Body

application/json
query
string
required
model_id
integer
required
conversation_id
string<uuid>
thread_id
string
max_tokens
integer
temperature
number
image_generation
boolean
image_parameters
object
persona
object
base_model
boolean
session_id
string
user_id
string
tenant_id
string
top_k
integer
top_p
number
response_length_preference
enum<string>
Available options:
concise,
balanced,
detailed
creativity_level
number
factual_precision
number

Response

Server-Sent Events stream of data: {...} chunks, terminated by data: [DONE]. The client assembles the text chunks into a single InferenceResult.

The generated response return signature envelope matching inference queries.

response
string
default:""

The core text answer generation string payload generated by the model.

conversation_id
string | null

The unique identification tracking string if part of a continued conversation.

thread_id
string | null

Thread tracking identifier reference index.

tokens_used
integer | null
latency_ms
number | null
metadata
Metadata · object | null