gigatoken 0.6.0

Tokenize your documents at GB/s

Gigatoken

~1000x faster than HuggingFace's tokenizers, drop-in replacement.

Tokenize your text data at GB/s!

Keep in mind that both HF tokenizers and tiktoken are already running multithreaded Rust!

What is Gigatoken?

Gigatoken is the fastest tokenizer for language modeling. It supports a wide range of CPU hardware, and nearly all commonly used tokenizers.

Installation

pip install gigatoken

Usage

Gigatoken can be used with its own API, or in compatibility mode with HuggingFace Tokenizers or Tiktoken.

Compatibility Mode (Easiest)

import gigatoken as gt

# Minimum change from existing HuggingFace tokenizers usage (compatibility mode)
hf_tokenizer = ...
tokenizer = gt.Tokenizer(hf_tokenizer).as_hf()

# tokenizer can be used in the same contexts as hf_tokenizer
tokens = tokenizer.encode_batch(["This is a test string", "And here is another"])

A substantial amount of effort has been put into making sure the outputs match exactly with what you would get with HuggingFace Tokenizers in this setting, but this is at a non-negligible cost to performance. You can still expect way faster performance across the board, but not quite the 1000x you will get with the Gigatoken API.

Gigatoken API (Fastest)

import gigatoken as gt

tokenizer = gt.Tokenizer("Qwen/Qwen3-8B")  # Accepts HF model names
file_source = gt.TextFileSource(["owt_train.txt"], separator=b"<|endoftext|>")
tokens = tokenizer.encode_files(file_source)

Using the Gigatoken API lets the Rust implementation read data directly, and skips as much overhead as possible while allowing for maximum parallelism. Keep in mind that passing Python data structures through this API still incurs the overhead of reading from Python.

FAQ

Q: Did you just way over-optimize for a specific CPU and tokenizer? How is it so fast?

No, I way over-optimized for every combination of these! The results are very consistent across CPUs (modern x86 and ARM), and across specific tokenizers.

The major improvements are in optimizing heavily an implementation that usually is outsourced to a Regex engine (pretokenization) using SIMD and other tricks, as well as heavily optimizing caching of pretoken mappings (if a word has been seen before, look it up its encoded tokens efficiently). In addition, interactions with Python are minimized, and threads are minimally interacting with each other.

Q: How can I quickly check if my tokenizer is supported?

You can try it out without installing anything! The following command will validate and time tokenization for a given HuggingFace model repo:

# Download your data
wget https://huggingface.co/datasets/stanford-cs336/owt-sample/resolve/main/owt_train.txt.gz  # Just an example!
gunzip owt_train.txt.gz

uvx --with tokenizers gigatoken bench 'openai-community/gpt2' owt_train.txt \
    --in-memory --validate --comparison-limit 100MB \
    --separator "<|endoftext|>"

gigatoken:    1.557 s |   11920.51 MB at  7653.89 MB/s |  2701.65 Mtok at 1734.67 Mtok/s
       hf:   15.268 s |     100.00 MB at     6.55 MB/s |    22.76 Mtok at    1.49 Mtok/s
gigatoken is 1168.58x faster than hf (by MB/s)
validation OK: 20401 documents match

This example uses the train sample from this dataset. You can see help for these flags with uvx gigatoken bench --help. Keep in mind that you might need to run twice on macOS to get a good reading since the first run will always perform a security scan.

Q: I've found a mismatch/slow use-case, is this expected?

Most likely not! Despite reasonably wide testing I don't have every use-case on hand, so please report anything you find in a GitHub Issue so I can address it as soon as possible.

---

AI Use Disclosure

A majority of this code base was crafted by hand without any use of AI (which can be seen from the project's Git history). In the final stages of the project, AI was used to assist:

• Implementing the user-facing API
• Widening of compatibility, for instance generalizing and porting the pretokenizer implementations to support more tokenizers, less interesting features like padding/truncation/unicode normalization
• Porting SIMD strategies between AVX512/AVX2/NEON
• Final profiling stages and the last ~4x worth of performance from eliminating branching and improving the pretoken cache hierarchy
• Refactoring and code reuse