axiom/dist/docs/evals/interfaces/EvalBuilder.md

Axiom AI SDK provides - an API to wrap your AI calls with observability instrumentation. - offline evals - online evals

[**axiom v0.51.1**](../../README.md)

***

[axiom](../../README.md) / [evals](../README.md) / EvalBuilder

# Interface: EvalBuilder\<AllowedFlags, TInput, TExpected, TOutput\>

## Type Parameters

### AllowedFlags

`AllowedFlags` *extends* `Record`\<`string`, `any`\> = \{ \}

### TInput

`TInput` *extends* `string` \| `Record`\<`string`, `any`\> = `string`

### TExpected

`TExpected` *extends* `string` \| `Record`\<`string`, `any`\> = `string`

### TOutput

`TOutput` *extends* `string` \| `Record`\<`string`, `any`\> = `string`

## Methods

### run()

> **run**(`suffix?`): `void`

#### Parameters

##### suffix?

`string`

#### Returns

`void`

***

### withFlags()

> **withFlags**\<`F`\>(`flags`): `EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>

#### Type Parameters

##### F

`F` *extends* `Partial`\<`AllowedFlags`\>

#### Parameters

##### flags

`F`

#### Returns

`EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>

***

### withModel()

> **withModel**(`model`): `EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>

#### Parameters

##### model

`string`

#### Returns

`EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>

***

### withTimeout()

> **withTimeout**(`timeout`): `EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>

#### Parameters

##### timeout

`number`

#### Returns

`EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>

***

### withTrials()

> **withTrials**(`trials`): `EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>

Set the number of times to run each case.
Each trial runs the task independently, and scores are aggregated per scorer.

#### Parameters

##### trials

`number`

#### Returns

`EvalBuilder`\<`AllowedFlags`, `TInput`, `TExpected`, `TOutput`\>