Crate tiktoken_wasm 

WebAssembly bindings for the tiktoken BPE tokenizer.

Provides browser-compatible wrappers around the core tiktoken crate, enabling high-performance token encoding, decoding, counting, and cost estimation directly in JavaScript/TypeScript applications.

All encoding instances are cached globally via OnceLock, so repeated calls to getEncoding() with the same name return the same underlying data.

Structs

Encoding

WASM wrapper around a tiktoken encoding instance.

ModelInfo

Model pricing and metadata.

Functions

all_models

List all supported models with pricing info.

encoding_for_model

Get an encoding for a model name (e.g. "gpt-4o", "o3-mini", "llama-4", "deepseek-r1").

estimate_cost

Estimate cost in USD for a given model, input token count, and output token count.

get_encoding

Get an encoding by name.

get_model_info

Get model pricing and metadata.

list_encodings

List all available encoding names.

model_to_encoding

Map a model name to its encoding name without loading the encoding.

models_by_provider

List models filtered by provider name.