Enum LlmError 

pub enum LlmError {
    ConfigError(String),
    ConnectionError(String),
    AuthenticationError,
    RateLimitError {
        retry_after: Option<u64>,
    },
    ModelNotFound {
        model: String,
    },
    ServerError {
        status: u16,
        message: String,
    },
    ParseError(String),
    OpenAiError(OpenAIError),
    ContentModeration {
        detail: String,
    },
    ContentModerationInput {
        detail: String,
    },
    ProviderError {
        code: String,
        message: String,
    },
}

Unified error type for LLM operations.

Variants

ConfigError(String)

ConnectionError(String)

AuthenticationError

RateLimitError

Fields

retry_after: Option<u64>

ModelNotFound

Fields

model: String

ServerError

Fields

status: u16

message: String

ParseError(String)

OpenAiError(OpenAIError)

ContentModeration

The provider’s content moderation blocked the model OUTPUT, pushed as an error event into the SSE stream (no choices field, so async-openai reports a JSON deserialization failure with the cause buried inside).

Fields

detail: String

ContentModerationInput

The provider’s content moderation blocked the INPUT (the request / conversation context was rejected with e.g. HTTP 400). Every subsequent turn resends the same context, so the session stays blocked until the context is cleared.

Fields

detail: String

ProviderError

Fields

code: String

message: String

Implementations

impl LlmError

pub fn from_openai_error(err: OpenAIError) -> Self

Map an async-openai error to a friendlier one where possible.

Two cases are improved:

• ApiError whose code/type is a known content-moderation code (e.g. DashScope data_inspection_failed) — the provider rejects the request outright with HTTP 400.
• JSONDeserialize where the raw payload is actually a provider error event {"error": {...}} pushed into the SSE stream — async-openai cannot parse it as a chunk (no choices), so the real cause would otherwise be buried in the deserialize error.

Everything else is passed through unchanged.

Trait Implementations

impl Debug for LlmError

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Display for LlmError

fn fmt(&self, __formatter: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Error for LlmError

fn source(&self) -> Option<&(dyn Error + 'static)>

Returns the lower-level source of this error, if any.

fn description(&self) -> &str

👎Deprecated since 1.42.0:

use the Display impl or to_string()

fn cause(&self) -> Option<&dyn Error>

👎Deprecated since 1.33.0:

replaced by Error::source, which can support downcasting

fn provide<'a>(&'a self, request: &mut Request<'a>)

🔬This is a nightly-only experimental API. (error_generic_member_access)

Provides type-based access to context intended for error reports.

impl From<OpenAIError> for LlmError

fn from(source: OpenAIError) -> Self

Converts to this type from the input type.