Struct RequestTiming 

pub struct RequestTiming {
    pub ttft_ms: f64,
    pub decode_ms: f64,
    pub output_tokens: u64,
}

2026-09-26: A finished request’s timing, as the scheduler reports it with the request’s terminal frame. A request that ends in an error frame reports none.

Fields

ttft_ms: f64

2026-09-26: Request start to decode start.

decode_ms: f64

2026-09-26: Decode start to the terminal frame.

output_tokens: u64

Implementations

impl RequestTiming

pub fn e2e_ns(&self) -> u64

2026-09-26: Request start to the terminal frame: ttft_ms + decode_ms, in ns.

pub fn tpot_ns(&self) -> Option<u64>

2026-09-26: Mean time per output token after the first; None for a request that produced fewer than two tokens.

Trait Implementations

impl Clone for RequestTiming

fn clone(&self) -> RequestTiming

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for RequestTiming

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

Formats the value using the given formatter.

impl PartialEq for RequestTiming

fn eq(&self, other: &RequestTiming) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Copy for RequestTiming

impl StructuralPartialEq for RequestTiming