struct OpenRouter::Reasoning

• OpenRouter::Reasoning
• Struct
• Value
• Object

Overview

Represents the reasoning configuration for models that support reasoning tokens

Included Modules

• JSON::Serializable

Defined in:

openrouter/types/reasoning.cr

Constructors

• .new(pull : JSON::PullParser)
• .new(effort : String | Nil = nil, max_tokens : Int32 | Nil = nil, exclude : Bool = false, enabled : Bool | Nil = nil)

Class Method Summary

• .enabled(exclude : Bool = false)

  Create a reasoning config with default parameters

• .high(exclude : Bool = false)

  Create a reasoning config with high effort

• .low(exclude : Bool = false)

  Create a reasoning config with low effort

• .medium(exclude : Bool = false)

  Create a reasoning config with medium effort

• .with_max_tokens(max_tokens : Int32, exclude : Bool = false)

  Create a reasoning config with specific max tokens

Instance Method Summary

• #effort : String | Nil

  Effort level for reasoning: "high", "medium", or "low" (OpenAI-style)

• #effort=(effort : String | Nil)

  Effort level for reasoning: "high", "medium", or "low" (OpenAI-style)

• #enabled : Bool | Nil

  Enable reasoning with default parameters (default: inferred from effort or max_tokens)

• #enabled=(enabled : Bool | Nil)

  Enable reasoning with default parameters (default: inferred from effort or max_tokens)

• #exclude : Bool

  Whether to exclude reasoning tokens from response (default: false)

• #exclude=(exclude : Bool)

  Whether to exclude reasoning tokens from response (default: false)

• #max_tokens : Int32 | Nil

  Specific token limit for reasoning (Anthropic-style)

• #max_tokens=(max_tokens : Int32 | Nil)

  Specific token limit for reasoning (Anthropic-style)

Constructor Detail

Class Method Detail

Instance Method Detail