Retry mechanism with exponential backoff for LLM API calls.
Configuration
config :elixir_llm, :retry,
max_attempts: 3,
base_delay_ms: 1000,
max_delay_ms: 30_000,
jitter: trueExample
# With default retry settings
ElixirLLM.new()
|> ElixirLLM.with_retry()
|> ElixirLLM.ask("Hello!")
# With custom retry settings
ElixirLLM.new()
|> ElixirLLM.with_retry(max_attempts: 5, base_delay_ms: 500)
|> ElixirLLM.ask("Hello!")Retry Behavior
By default, the following errors are retried:
- Rate limit errors (429)
- Server errors (5xx)
- Network errors
- Timeout errors
Non-retryable errors (authentication, validation) are returned immediately.
Summary
Functions
Calculates the delay for a given retry attempt using exponential backoff.
Returns true if the error should trigger a retry.
Executes a function with retry logic.
Types
@type retry_opts() :: [ max_attempts: pos_integer(), base_delay_ms: pos_integer(), max_delay_ms: pos_integer(), jitter: boolean(), on_retry: (integer(), term() -> any()) ]
Functions
@spec calculate_delay(pos_integer(), map()) :: non_neg_integer()
Calculates the delay for a given retry attempt using exponential backoff.
Returns true if the error should trigger a retry.
@spec with_retry((-> {:ok, term()} | {:error, term()}), retry_opts()) :: {:ok, term()} | {:error, term()}
Executes a function with retry logic.
Options
:max_attempts- Maximum number of attempts (default: 3):base_delay_ms- Base delay between retries in milliseconds (default: 1000):max_delay_ms- Maximum delay between retries (default: 30000):jitter- Add random jitter to delays (default: true):on_retry- Callback called before each retry:(attempt, error) -> any()
Example
ElixirLLM.Retry.with_retry(fn ->
provider.chat(chat)
end, max_attempts: 5)