Struct TokenUsage 

pub struct TokenUsage {
    pub prompt_tokens: u32,
    pub cached_input_tokens: Option<u32>,
    pub cache_write_tokens: Option<u32>,
    pub completion_tokens: u32,
    pub total_tokens: u32,
    pub tool_call_count: Option<u32>,
    pub prompt_eval_nanos: u64,
    pub eval_nanos: u64,
    pub total_nanos: u64,
    pub load_nanos: u64,
    pub advisor: Option<AdvisorTokenUsage>,
}

Token usage statistics from the LLM.

Fields

prompt_tokens: u32

Number of tokens in the prompt (input).

cached_input_tokens: Option<u32>

Input tokens served from cache, when the provider reports them.

cache_write_tokens: Option<u32>

Input tokens written to the provider’s prompt cache, when the provider reports them. Written input is priced apart from fresh input and from cache reads, so a spend figure needs it. These tokens are part of prompt_tokens.

completion_tokens: u32

Number of tokens generated by the model (output).

total_tokens: u32

Total tokens consumed (prompt + completion).

tool_call_count: Option<u32>

Tool/function calls consumed during the request, when reported.

prompt_eval_nanos: u64

Time to evaluate the prompt (nanoseconds). Ollama: prompt_eval_duration.

eval_nanos: u64

Time to generate completion tokens (nanoseconds). Ollama: eval_duration.

total_nanos: u64

Total wall-clock time for the request (nanoseconds). Ollama: total_duration.

load_nanos: u64

Time spent loading the model (nanoseconds). Ollama: load_duration.

advisor: Option<AdvisorTokenUsage>

What a second model, offered to the answering one as an advisor, spent on this request. It is billed at the advisor’s rates, so it is not in the counts above. None where no advisor was offered or the provider has none.

Trait Implementations

impl Clone for TokenUsage

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TokenUsage

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for TokenUsage

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for TokenUsage

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.