Skip to main content
GET
Benchmark Trend

Authorizations

Authorization
string
header
required

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

Query Parameters

benchmark
string
required

Benchmark name, gsm8k for example

metric
string
required

Metric key, exact_match,strict-match for example

project_key
string
default:""

Filter by project; empty = everything visible

model
string
default:""

Compute the variance across repeated evaluations of this model

Response

Successful Response

benchmark
string
required
metric
string
required
points
BenchmarkTrendPointOut · object[]
required
variance
BenchmarkVarianceOut · object | null
required

How far repeated evaluations of one model spread (sample standard deviation).