AI News

Trending story

Georgi Gerganov announces the release of llama.cpp v1.0.0

5 reports5 discussions10 story updatesUpdated Nov 14 21:37

Understand this story

AI overview

On November 13, 2026, llama.cpp author Georgi Gerganov posted a release link on Twitter and announced the release of llama.cpp v1.0.0. On the same day, Hugging Face CEO Clément Delangue and Gerganov jointly announced the release of llama.cpp v1, calling it the first fully stable major version of the most widely used library for running AI models locally, capable of running on phones, laptops, and workstations. Delangue said it lets users own their own intelligence for free, without having to send data to others, and reflects a commitment to supporting all kinds of hardware, platforms, and model types. The version has now been released, and there is no further development information at present.

AI-generated from public reports

Latest development · Report

llama.cpp and the ggml library reach version 1.0.0, with explicit API compatibility rules

Developments

1 confirmed developments

  1. Georgi Gerganov announces the release of llama.cpp v1.0.0

    Source:Clément Delangue

Reports and discussion timeline

Browse reports and discussion by time. Their evidence levels remain distinct.

Nov 14

  1. IT Home · Artificial Intelligence

    llama.cpp and the ggml library reach version 1.0.0, with explicit API compatibility rules

    Developer Georgi Gerganov announced that the open-source inference project llama.cpp and the underlying machine learning engine ggml have reached library version 1.0.0, and clarified that major version numbers correspond to API or ABI incompatible changes, minor version numbers correspond to backward-compatible features and model support, and patch version numbers correspond to bug fixes that do not change the API. llama.cpp already supports more than 150 model architecture variants and mainstream speculative decoding methods, and the team has also used Hugging Face's HF Buckets and HF Jobs to scale CI, while expanding self-hosted testing coverage with support from NVIDIA, AMD, Qualcomm, Arm, IBM, and Intel hardware.

  2. IT Home · AIRelated discussion

    IT Home · AI joined this discussion

    This post contributes to heat; it is not a formal report or confirmed fact.

  3. @mark_kRelated discussion

    @mark_k joined this discussion

    This post contributes to heat; it is not a formal report or confirmed fact.

  4. Merve NoyanRelated discussion

    Merve Noyan joined this discussion

    This post contributes to heat; it is not a formal report or confirmed fact.

  5. gihyo.jpRelated discussion

    gihyo.jp joined this discussion

    This post contributes to heat; it is not a formal report or confirmed fact.

  6. Daniel HanRelated discussion

    Daniel Han joined this discussion

    This post contributes to heat; it is not a formal report or confirmed fact.

  7. Thomas Wolf

    llama.cpp releases v1.0.0

    llama.cpp releases v1.0.0; the author Georgi Gerganov announced this version. When reposting, Thomas Wolf said this is an important moment in the library's lifecycle and mentioned that transformers released version 1.0 in 2019.

Nov 13

  1. Clément DelangueSelected

    Hugging Face CEO announces llama.cpp v1 release

    Hugging Face CEO Clément Delangue and llama.cpp author Georgi Gerganov announced the release of llama.cpp v1, the first fully stable major version of the most widely used library for running AI models locally, capable of running on phones, laptops, and workstations. Delangue said it lets users own their own intelligence for free, without sending data to others, and reflects a commitment to supporting all kinds of hardware, platforms, and model types.

  2. Unite.AI

    llama.cpp and ggml Release v1.0.0, Establishing API Compatibility Versioning Rules

    ggml-org released llama.cpp v1.0.0 on November 13, and the ggml tensor library also reached v1.0.0; going forward, the two projects will signal source-level API and ABI incompatibilities through major version number changes. This version removed 35 deprecated public API functions, promoted the NextN speculative decoding API from llama-ext.h to the public llama.h, and added inference capabilities such as FP8 KV cache and host memory caching for MoE models. llama-server fixed SSRF, CORS, and DNS rebinding issues, added an OpenAI-compatible /v1/audio/speech endpoint, and official release tags are now created only when the internal ggml copy matches the released ggml.

  3. Georgi GerganovSelected

    llama.cpp releases v1.0.0

    llama.cpp v1.0.0 is now released, and author Georgi Gerganov provided the release link in a tweet.

Heat trend for this story

Current
23
Period peak
75 Nov 14 01:00
24-hour change
-67%

Move the pointer or select the chart to inspect hourly heat. Use the left and right arrow keys with a keyboard.

Related stories

  1. Related storiesEnglish translation is being prepared.
  2. Related storiesEnglish translation is being prepared.
  3. Related storiesUnsloth releases MiniMax-M3.1-Flash 2-bit GGUF
  4. Related storiesMiniMax releases the H3-Image image model and opens the weights