starlet/anthropic

Anthropic provider for starlet.

Uses the Anthropic Messages API for chat completions with Claude models.

Usage

import starlet
import starlet/anthropic

let client = anthropic.new(api_key)

starlet.chat(client, "claude-haiku-4-5-20251001")
|> starlet.user("Hello!")
|> starlet.send()

Extended Thinking

For models that support extended thinking, configure a thinking budget:

starlet.chat(client, "claude-haiku-4-5-20251001")
|> anthropic.with_thinking(16384)
|> starlet.max_tokens(32000)
|> starlet.user("Analyze this complex problem...")
|> starlet.send()

Note on max_tokens

Anthropic requires max_tokens in every request. If not explicitly set via starlet.max_tokens(), a default of 4096 is used.

Values

new

</>

pub fn new(api_key: String) -> starlet.Client(@internal Ext)

Creates a new Anthropic client with the given API key. Uses the default base URL: https://api.anthropic.com

Note: Anthropic requires max_tokens. If not explicitly set via starlet.max_tokens(), a default of 4096 is used.

new_with_base_url

</>

pub fn new_with_base_url(
  api_key: String,
  base_url: String,
) -> starlet.Client(@internal Ext)

Creates a new Anthropic client with a custom base URL. Useful for proxies or self-hosted endpoints.

Note: Anthropic requires max_tokens. If not explicitly set via starlet.max_tokens(), a default of 4096 is used.

thinking

</>

pub fn thinking(
  turn: starlet.Turn(t, f, @internal Ext),
) -> option.Option(String)

Get the thinking content from an Anthropic turn (if present).

with_thinking

</>

pub fn with_thinking(
  chat: starlet.Chat(t, f, s, @internal Ext),
  budget: Int,
) -> Result(
  starlet.Chat(t, f, s, @internal Ext),
  starlet.StarletError,
)

Enable extended thinking with a token budget. Budget must be at least 1024 tokens. The upper bound (less than max_tokens) is enforced by the API at request time.