[Doc] Clarify fastokens availability (#45813)
Signed-off-by: LjjJzd <3542531707@qq.com> Co-authored-by: OpenAI Codex <codex@openai.com>
LiJzd committed
Jul 6, 2026 at 21:33 UTC
641cb595928a8914a7020e60125ed94496421801
1 file changed
+3
-2
docs/configuration/optimization.md
+3
-2
@@ -276,8 +276,9 @@ By default vLLM uses the standard Hugging Face `tokenizers` library to power
276
the fast tokenizer. For BPE tokenizers (Qwen, Llama, DeepSeek, GPT-OSS, etc.)
277
you can switch to the [fastokens](https://github.com/crusoecloud/fastokens)
278
Rust backend, a drop-in replacement that's substantially faster on
279
-encode/decode and on streaming detokenization. Enable it by setting
280
-`VLLM_USE_FASTOKENS=1`:
279
+encode/decode and on streaming detokenization. `VLLM_USE_FASTOKENS` is
280
+available in vLLM v0.23.0 and later. If your installed vLLM version does not
281
+recognize the environment variable, upgrade vLLM before enabling the override:
282
283
```console
284
VLLM_USE_FASTOKENS=1 vllm serve Qwen/Qwen3-8B