prompt-executor-anthropic-client/ai.koog.prompt.executor.clients.anthropic/AnthropicParams

AnthropicParams

class AnthropicParams(temperature: Double? = null, maxTokens: Int? = null, numberOfChoices: Int? = null, speculation: String? = null, schema: LLMParams.Schema? = null, toolChoice: LLMParams.ToolChoice? = null, user: String? = null, additionalProperties: Map<String, JsonElement>? = null, val topP: Double? = null, val topK: Int? = null, val stopSequences: List<String>? = null, val container: String? = null, val mcpServers: List<AnthropicMCPServerURLDefinition>? = null, val serviceTier: AnthropicServiceTier? = null, val thinking: AnthropicThinking? = null, val cacheControl: AnthropicCacheControl? = null) : LLMParams(source)

Anthropic Messages API parameters layered on top of LLMParams.

Constructors

AnthropicParams

constructor(temperature: Double? = null, maxTokens: Int? = null, numberOfChoices: Int? = null, speculation: String? = null, schema: LLMParams.Schema? = null, toolChoice: LLMParams.ToolChoice? = null, user: String? = null, additionalProperties: Map<String, JsonElement>? = null, topP: Double? = null, topK: Int? = null, stopSequences: List<String>? = null, container: String? = null, mcpServers: List<AnthropicMCPServerURLDefinition>? = null, serviceTier: AnthropicServiceTier? = null, thinking: AnthropicThinking? = null, cacheControl: AnthropicCacheControl? = null)

Properties

additionalProperties

val additionalProperties: Map<String, JsonElement>?

cacheControl

val cacheControl: AnthropicCacheControl?

when automatic cache control provided, AnthropicCacheControl is added at the request top level and the API automatically applies the cache breakpoint to the last cacheable block. This is the recommended approach for multi-turn conversations where you don't need fine-grained control over breakpoint positions.

container

val container: String?

Container identifier for reuse across requests.

maxTokens

val maxTokens: Int?

mcpServers

val mcpServers: List<AnthropicMCPServerURLDefinition>?

MCP servers to be used in this request

numberOfChoices

val numberOfChoices: Int?

schema

val schema: LLMParams.Schema?

serviceTier

val serviceTier: AnthropicServiceTier?

Determines whether to use priority capacity (if available) or standard capacity for this request.

speculation

val speculation: String?

stopSequences

val stopSequences: List<String>?

Custom text sequences that cause the model to stop generating. If matched, response will have stop_reason of "stop_sequence".

temperature

val temperature: Double?

thinking

val thinking: AnthropicThinking?

Configuration for enabling Claude's extended thinking.

toolChoice

val toolChoice: LLMParams.ToolChoice?

topK

val topK: Int?

Sample only from top K options for each subsequent token (≥ 0). Recommended for advanced use cases.

topP

val topP: Double?

Nucleus sampling in (0.0, 1.0]; use instead of temperature. Cuts off token sampling at a cumulative probability threshold.

user

val user: String?

Functions

component1

operator fun component1(): Double?

component10

operator fun component10(): Map<String, JsonElement>?

component2

operator fun component2(): Int?

component3

operator fun component3(): Int?

component4

operator fun component4(): String?

component5

operator fun component5(): LLMParams.Schema?

component6

operator fun component6(): LLMParams.ToolChoice?

component7

operator fun component7(): String?

copy

open override fun copy(temperature: Double?, maxTokens: Int?, numberOfChoices: Int?, speculation: String?, schema: LLMParams.Schema?, toolChoice: LLMParams.ToolChoice?, user: String?, additionalProperties: Map<String, JsonElement>?): AnthropicParams

fun copy(temperature: Double? = this.temperature, maxTokens: Int? = this.maxTokens, numberOfChoices: Int? = this.numberOfChoices, speculation: String? = this.speculation, schema: LLMParams.Schema? = this.schema, toolChoice: LLMParams.ToolChoice? = this.toolChoice, user: String? = this.user, additionalProperties: Map<String, JsonElement>? = this.additionalProperties, topP: Double? = this.topP, topK: Int? = this.topK, stopSequences: List<String>? = this.stopSequences, container: String? = this.container, mcpServers: List<AnthropicMCPServerURLDefinition>? = this.mcpServers, serviceTier: AnthropicServiceTier? = this.serviceTier, thinking: AnthropicThinking? = this.thinking, cacheControl: AnthropicCacheControl? = this.cacheControl): AnthropicParams

Creates a copy of this instance with the ability to modify any of its properties.

default

fun default(default: LLMParams): LLMParams

equals

open operator override fun equals(other: Any?): Boolean

hashCode

open override fun hashCode(): Int

toString

open override fun toString(): String