tokenizers Documentation

Fast State-of-the-Art Tokenizers optimized for Research and Production

10,947 stars Rust Rust 1 file · ~2,895 tokens #bert#gpt#language-model#natural-language-processing#natural-language-understanding#nlp#transformers
RAW Doc

File: README.md

Provides an implementation of today's most used tokenizers, with a focus on performance and
versatility.

Main features:

  • Train new vocabularies and tokenize, using today's most used tokenizers.
  • Extremely fast (both training and tokenization), thanks to the Rust implementation. Takes
    less than 20 seconds to tokenize a GB of text on a server's CPU.
  • Easy to use, but also extremely versatile.
  • Designed for research and production.
  • Normalization comes with alignments tracking. It's always possible to get the part of the
    original sentence that corresponds to a given token.
  • Does all the pre-processing: Truncate, Pad, add the special tokens your model needs.

Performances

Performances can vary depending on hardware, but running the ~/bindings/python/benches/test_tiktoken.py should give the following on a g6 aws instance:

Bindings

We provide bindings to the following languages (more to come!):

Installation

You can install from source using:

bash
pip install git+https://github.com/huggingface/tokenizers.git#subdirectory=bindings/python

or install the released versions with

bash
pip install tokenizers

Quick example using Python:

Choose your model between Byte-Pair Encoding, WordPiece or Unigram and instantiate a tokenizer:

python
from tokenizers import Tokenizer
from tokenizers.models import BPE

tokenizer = Tokenizer(BPE())

You can customize how pre-tokenization (e.g., splitting into words) is done:

python
from tokenizers.pre_tokenizers import Whitespace

tokenizer.pre_tokenizer = Whitespace()

Then training your tokenizer on a set of files just takes two lines of codes:

python
from tokenizers.trainers import BpeTrainer

trainer = BpeTrainer(special_tokens=["[UNK]", "[CLS]", "[SEP]", "[PAD]", "[MASK]"])
tokenizer.train(files=["wiki.train.raw", "wiki.valid.raw", "wiki.test.raw"], trainer=trainer)

Once your tokenizer is trained, encode any text with just one line:

python
output = tokenizer.encode("Hello, y'all! How are you 😁 ?")
print(output.tokens)
# ["Hello", ",", "y", "'", "all", "!", "How", "are", "you", "[UNK]", "?"]

Check the documentation
or the quicktour to learn more!

File: bindings/python/README.md

Tokenizers

Provides an implementation of today's most used tokenizers, with a focus on performance and
versatility.

Bindings over the Rust implementation.
If you are interested in the High-level design, you can go check it there.

Otherwise, let's dive in!

Main features:

  • Train new vocabularies and tokenize using 4 pre-made tokenizers (Bert WordPiece and the 3
    most common BPE versions).
  • Extremely fast (both training and tokenization), thanks to the Rust implementation. Takes
    less than 20 seconds to tokenize a GB of text on a server's CPU.
  • Easy to use, but also extremely versatile.
  • Designed for research and production.
  • Normalization comes with alignments tracking. It's always possible to get the part of the
    original sentence that corresponds to a given token.
  • Does all the pre-processing: Truncate, Pad, add the special tokens your model needs.

Installation

With pip:

bash
pip install tokenizers

From sources:

To use this method, you need to have the Rust installed:

bash
# Install with:
curl https://sh.rustup.rs -sSf | sh -s -- -y
export PATH="$HOME/.cargo/bin:$PATH"

Once Rust is installed, you can compile doing the following

bash
git clone https://github.com/huggingface/tokenizers
cd tokenizers/bindings/python

# Create a virtual env (you can use yours as well)
python -m venv .env
source .env/bin/activate

# Install `tokenizers` in the current virtual env
pip install -e .

Free-threaded Python (3.14t)

tokenizers ships dedicated wheels for the free-threaded build of CPython
(python3.14t). These wheels declare Py_MOD_GIL_NOT_USED, so importing
tokenizers does not force the GIL back on — multi-threaded code stays
GIL-free.

The full mutable API works on 3.14t — the same as on regular CPython.
Setters are thread-safe: the inner tokenizer state is wrapped in a
std::sync::RwLock, so concurrent tokenizer.X = … from multiple threads
serialize correctly and concurrent encode operations take a read guard
that blocks writers only briefly.

python
from tokenizers import Tokenizer
from tokenizers.models import BPE
from tokenizers.pre_tokenizers import Whitespace
from tokenizers.processors import ByteLevel

tok = Tokenizer(BPE())
tok.pre_tokenizer = Whitespace()                 # ✅ thread-safe on 3.14t
tok.post_processor = ByteLevel(trim_offsets=True)

Caveat — compound mutations are not atomic. Statements like
tokenizer.post_processor.special_tokens = X evaluate in two steps from
Python's point of view (read attribute → set attribute on the result). If
another thread swaps tokenizer.post_processor between those steps, the
mutation lands on an orphaned component. This is the same class of race
as dict[k] = v interleaved with dict.clear() — coordinate with a Python
lock if you need the compound to be atomic.

For the full thread-safety analysis, see
docs/free-threading-audit.md.

Load a pretrained tokenizer from the Hub

python
from tokenizers import Tokenizer

tokenizer = Tokenizer.from_pretrained("bert-base-cased")

Using the provided Tokenizers

We provide some pre-build tokenizers to cover the most common cases. You can easily load one of
these using some vocab.json and merges.txt files:

python
from tokenizers import CharBPETokenizer

# Initialize a tokenizer
vocab = "./path/to/vocab.json"
merges = "./path/to/merges.txt"
tokenizer = CharBPETokenizer(vocab, merges)

# And then encode:
encoded = tokenizer.encode("I can feel the magic, can you?")
print(encoded.ids)
print(encoded.tokens)

And you can train them just as simply:

python
from tokenizers import CharBPETokenizer

# Initialize a tokenizer
tokenizer = CharBPETokenizer()

# Then train it!
tokenizer.train([ "./path/to/files/1.txt", "./path/to/files/2.txt" ])

# Now, let's use it:
encoded = tokenizer.encode("I can feel the magic, can you?")

# And finally save it somewhere
tokenizer.save("./path/to/directory/my-bpe.tokenizer.json")

Provided Tokenizers

  • CharBPETokenizer: The original BPE
  • ByteLevelBPETokenizer: The byte level version of the BPE
  • SentencePieceBPETokenizer: A BPE implementation compatible with the one used by SentencePiece
  • BertWordPieceTokenizer: The famous Bert tokenizer, using WordPiece

All of these can be used and trained as explained above!

Build your own

Whenever these provided tokenizers don't give you enough freedom, you can build your own tokenizer,
by putting all the different parts you need together.
You can check how we implemented the provided tokenizers and adapt them easily to your own needs.

Building a byte-level BPE

Here is an example showing how to build your own byte-level BPE by putting all the different pieces
together, and then saving it to a single file:

python
from tokenizers import Tokenizer, models, pre_tokenizers, decoders, trainers, processors

# Initialize a tokenizer
tokenizer = Tokenizer(models.BPE())

# Customize pre-tokenization and decoding
tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel(add_prefix_space=True)
tokenizer.decoder = decoders.ByteLevel()
tokenizer.post_processor = processors.ByteLevel(trim_offsets=True)

# And then train
trainer = trainers.BpeTrainer(
    vocab_size=20000,
    min_frequency=2,
    initial_alphabet=pre_tokenizers.ByteLevel.alphabet()
)
tokenizer.train([
    "./path/to/dataset/1.txt",
    "./path/to/dataset/2.txt",
    "./path/to/dataset/3.txt"
], trainer=trainer)

# And Save it
tokenizer.save("byte-level-bpe.tokenizer.json", pretty=True)

Now, when you want to use this tokenizer, this is as simple as:

python
from tokenizers import Tokenizer

tokenizer = Tokenizer.from_file("byte-level-bpe.tokenizer.json")

encoded = tokenizer.encode("I can feel the magic, can you?")

Typing support and stub generation

The compiled PyO3 extension does not expose type annotations, so editors and type checkers would otherwise see most objects as Any. To provide full typing support, we use a two-step stub generation process:

  1. Rust introspection (tools/stub-gen/): Uses pyo3-introspection to analyze the compiled extension and generate .pyi stub files
  2. Python enrichment (stub.py): Adds docstrings from the runtime module and generates forwarding __init__.py shims

Running stub generation

The easiest way to regenerate stubs is via make style:

bash
cd bindings/python
make style

This will:

  1. Build the extension with maturin develop --release
  2. Run introspection to generate .pyi files
  3. Enrich stubs with docstrings via stub.py
  4. Format with ruff

Running manually

To run the stub generator directly:

bash
cd bindings/python
cargo run --manifest-path tools/stub-gen/Cargo.toml
python stub.py

The stub generator automatically:

  • Builds the extension using maturin
  • Copies the built .so to the project root for introspection
  • Detects and sets PYTHONHOME for embedded Python (handles uv/venv environments)
  • Generates stubs to py_src/tokenizers/

Troubleshooting

If you encounter Python initialization errors, you can manually set PYTHONHOME:

bash
export PYTHONHOME=$(python3 -c 'import sys; print(sys.base_prefix)')
cargo run --manifest-path tools/stub-gen/Cargo.toml

File: bindings/node/npm/win32-x64-msvc/README.md

`tokenizers-win32-x64-msvc`

This is the x86_64-pc-windows-msvc binary for tokenizers

File: bindings/node/npm/win32-ia32-msvc/README.md

`tokenizers-win32-ia32-msvc`

This is the i686-pc-windows-msvc binary for tokenizers

File: bindings/node/npm/win32-arm64-msvc/README.md

`tokenizers-win32-arm64-msvc`

This is the aarch64-pc-windows-msvc binary for tokenizers

File: bindings/node/npm/linux-x64-musl/README.md

`tokenizers-linux-x64-musl`

This is the x86_64-unknown-linux-musl binary for tokenizers

File: bindings/node/npm/linux-x64-gnu/README.md

`tokenizers-linux-x64-gnu`

This is the x86_64-unknown-linux-gnu binary for tokenizers

File: bindings/node/npm/linux-arm64-musl/README.md

`tokenizers-linux-arm64-musl`

This is the aarch64-unknown-linux-musl binary for tokenizers

File: bindings/node/npm/linux-arm64-gnu/README.md

`tokenizers-linux-arm64-gnu`

This is the aarch64-unknown-linux-gnu binary for tokenizers

File: bindings/node/npm/linux-arm-gnueabihf/README.md

`tokenizers-linux-arm-gnueabihf`

This is the armv7-unknown-linux-gnueabihf binary for tokenizers

Explore Related Technologies & Topics

View All 9,111+ Repositories →

Discover similar high-velocity repositories, agent skills, and OpenAPI specifications across the ecosystem.

#bert Browse bert catalog → #gpt Browse gpt catalog → #language-model Browse language-model catalog → #natural-language-processing Browse natural-language-processing catalog → #natural-language-understanding Browse natural-language-understanding catalog → #nlp Browse nlp catalog → ⚡ MCP Servers Model Context Protocol → 🧠 Agent Skills Prompt & rule files → 🔌 OpenAPI Specs REST & Swagger schemas →