LlmProvider

LlmProvider selects the backend client Appam will instantiate at runtime.

Definition

pub enum LlmProvider {
    OpenRouterCompletions,
    OpenRouterResponses,
    Anthropic,
    OpenAI,
    OpenAICodex,
    Vertex,
    AzureOpenAI { resource_name: String, api_version: String },
    AzureAnthropic { base_url: String, auth_method: AzureAnthropicAuthMethod },
    Bedrock { region: String, model_id: String, auth_method: BedrockAuthMethod },
}

Default is OpenRouterCompletions.

Parsing and display

FromStr accepts:

• openrouter, openrouter-completions, openroutercompletions
• openrouter-responses, openrouterresponses
• anthropic
• openai
• openai-codex, openai_codex, codex
• vertex, google-vertex, google_vertex
• azure-openai, azure_openai, azure
• azure-anthropic, azure_anthropic
• bedrock, aws-bedrock, aws_bedrock

When parsing azure* or bedrock, extra fields are filled from environment variables inside from_str(...).

Display returns:

• openrouter-completions
• openrouter-responses
• anthropic
• openai
• openai-codex
• vertex
• azure-openai[resource@api-version]
• azure-anthropic[base-url:auth]
• bedrock[model_id@region:auth]

Pricing normalization

pricing_key() intentionally collapses variants:

Example

use appam::prelude::*;

let agent = AgentBuilder::new("assistant")
    .provider(LlmProvider::OpenAI)
    .model("gpt-5.4")
    .system_prompt("You are helpful.")
    .build()?;