# li-plus/chatglm.cpp

C++ implementation of ChatGLM-6B & ChatGLM2-6B & ChatGLM3 & GLM4(V)

Repository: https://github.com/li-plus/chatglm.cpp
Canonical: https://ross.abutalabs.com/products/chatglmcpp
Language: C++
License: MIT
License Family: permissive
Topics: large-language-models, chatglm, nlp, chatglm2, chatglm3, codegeex2-6b, glm4
Last push: 2024-07-31T06:12:02+00:00

## Health v2 (maintenance only)
Score: 20/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 85
- inputs: {"age_days": 1198, "days_push": 763, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2962, forks 326 (observed 2026-08-28T04:07:32.378136+00:00)

## What it is
A pure C++ implementation of ChatGLM-6B, ChatGLM2-6B, ChatGLM3, and GLM-4(V) based on ggml, enabling memory-efficient real-time chatting on consumer hardware. It provides int4/int8 quantization, streaming generation, and Python bindings with web demo and API server options.

## Use cases
- run chatglm locally on a macbook
- quantize chatglm-6b to int4 for cpu inference
- serve a chatglm model behind an api
- chat with glm-4 offline without a gpu
- embed chatglm inference in a python app
- run codegeex2 for code completion locally

## When to choose
- you want fast, memory-efficient local inference of ChatGLM/GLM-4 models on CPU or Apple Silicon
- you need llama.cpp-style quantized GGML models with streaming output
- you want Python bindings plus CLI and API server options in one project

## When to avoid
- you need models outside the ChatGLM/GLM/CodeGeeX family
- you require heavy GPU multi-node inference or production-scale serving
- you need fine-grained training or finetuning rather than inference

## Facets
- artifact type: library
- maturity: active
- function: llm-inference, machine-learning, chatbot, cli, http-server
- domain: large-language-models, machine-learning, developer-tools
- platform: windows, python, cpp, cli
- tags: ggml, quantization, chatglm, glm4, cpu-inference, llama.cpp-style, local-llm, python-bindings, natural-language-processing, linux, macos

## Member repositories
- li-plus/chatglm.cpp (main) score 20

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:32.378136+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:32:08.654192+00:00, confidence not recorded.
  - readme: https://github.com/li-plus/chatglm.cpp (fetched 2026-08-28T04:07:32.378136+00:00, sha 0e9930349290)
- Data as of 2026-08-30T08:39:29.467469+00:00.