# tinker_cookbook.tokenizer_utils.get_tokenizer

### [**tinker_cookbook.tokenizer_utils.get_tokenizer**](https://github.com/thinking-machines-lab/tinker-cookbook/blob/main/tinker_cookbook/tokenizer_utils.py#L186)( _model_name_)

Get a tokenizer by name.

Checks the custom registry first (see [`register_tokenizer`](https://tinker-docs.thinkingmachines.ai/cookbook/api-reference/tokenizer_utils/register_tokenizer/)), then returns a tml-renderers tokenizer for Inkling, or falls back to HuggingFace `AutoTokenizer`. HuggingFace tokenizers are cached after first load.

**Parameters:**

- [**model_name**](https://github.com/thinking-machines-lab/tinker-cookbook/blob/main/tinker_cookbook/tokenizer_utils.py#L186) ( _str_) – HuggingFace model identifier (e.g. "Qwen/Qwen3-8B") or a custom registered name.

**Returns:** _Tokenizer_ – A tml-renderers tokenizer for Inkling, otherwise a HuggingFace `PreTrainedTokenizer`.

```
from tinker_cookbook.tokenizer_utils import get_tokenizer
tokenizer = get_tokenizer("Qwen/Qwen3-8B")
tokens = tokenizer.encode("Hello world")
```

## Referenced by

- [tinker_cookbook.tokenizer_utils.register_tokenizer](https://tinker-docs.thinkingmachines.ai/cookbook/api-reference/tokenizer_utils/register_tokenizer/)