# llama.cpp release b11813 adds Tekken tokenizer and new binaries

> The open‑source Llama.cpp library released version b11813 on Oct 22, 2026, swapping its default tokenizer for a Tekken‑based one and publishing fresh macOS, iOS, Linux and GPU binaries.

Oossa · 2026-10-22 · https://oossa.com/en/llama-cpp-release-b11813-adds-tekken-tokenizer-and-new-binaries

The ggml‑org team pushed llama.cpp version b11813 on Oct 22, 2026. The update replaces the Llama 3 tokenizer with a Tekken pre‑tokenizer for the pixtral model. It also bundles new pre‑built binaries for macOS (Apple Silicon and Intel), iOS, and several Ubuntu builds, including Vulkan and CUDA GPU versions.

## The facts

- Release tag b11813 published on Thu Oct 22 2026
- Adds Tekken pre‑tokenizer for pixtral model

## Why it matters

Developers can now run Llama.cpp with the new tokenizer and ready‑to‑use binaries on more platforms, lowering setup time.

## Sources & references

1. [ggml-org/llama.cpp b11813](https://github.com/ggml-org/llama.cpp/releases/tag/b11813) – llama.cpp, 2026-10-22

Last updated: 2026-10-22