explaingit

nichtdax/awesome-totally-open-chatgpt

4,769Audience · developerComplexity · 1/5Setup · easy

TLDR

A curated list of open-source alternatives to ChatGPT, tagged by completeness, bare code only, standard with model weights, or full with a working chat interface, covering only models specifically trained for conversation.

Mindmap

mindmap
  root((Open ChatGPT Alts))
    Completeness Tags
      Bare code only
      Standard with weights
      Full with interface
      Complicated caveats
    Use Cases
      Self-hosted chatbots
      Roleplay models
      Non-English models
      Consumer hardware
    Criteria
      No OpenAI dependency
      Instruction-tuned only
      Open weights required
    Audience
      Developers
      Researchers
      Privacy-focused users
Click or tap to explore — scroll the page freely

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

Things people build with this

USE CASE 1

Find and compare open-source chatbot projects to self-host instead of relying on ChatGPT

USE CASE 2

Identify which projects include ready-to-use model weights and a chat interface versus bare source code

USE CASE 3

Discover open-source models optimized for consumer hardware with limited memory

USE CASE 4

Find alternatives designed for non-English languages or specific use cases like creative writing

Getting it running

Difficulty · easy Time to first run · 5min

This is a reference list, not software. Setup complexity depends entirely on which project you choose from it.

In plain English

This repository is a curated list of open-source alternatives to ChatGPT. ChatGPT is a chatbot built by OpenAI on top of a language model that was trained with human feedback to follow instructions and hold conversations. Because ChatGPT itself is a closed commercial product, many researchers and developers have published their own versions using openly available models and training methods. This list collects those projects in one place. The list applies a clear standard for what counts as an alternative: the project must use a language model that has been specifically trained or fine-tuned to respond to instructions and chat, not just a raw text-completion model. Projects that simply wrap OpenAI's own APIs are excluded, because they depend on the same closed system the list is meant to offer an alternative to. Each entry is tagged to describe how complete it is. A "Bare" project is just source code with no prepackaged model weights or ready-to-use interface. A "Standard" project includes both the code and the trained model weights with basic API access. A "Full" project adds a usable chat interface, such as a web page or desktop application. A "Complicated" tag is used for projects that have important caveats, such as depending partly on a closed or restricted model. The projects listed span a range of approaches and use cases. Some focus on reproducing the RLHF training process that makes a model conversational. Others prioritize running a capable model on consumer hardware with limited memory. Several are designed for creative writing and roleplay rather than general assistance. A few target non-English languages. This is a community reference document rather than a piece of software. It is most useful for someone who wants to find, compare, or choose from among the many open alternatives to closed AI chat systems.

Copy-paste prompts

Prompt 1
Using the awesome-totally-open-chatgpt list, which Full-tagged projects have a working web chat interface I can run on a machine with 8GB RAM today?
Prompt 2
I want to self-host an open-source chatbot for my team. Based on the awesome-totally-open-chatgpt list, which Standard or Full projects have the most active communities?
Prompt 3
Which projects in awesome-totally-open-chatgpt support languages other than English and include a deployable web interface?
Prompt 4
Compare the RLHF-trained projects in awesome-totally-open-chatgpt, which ones reproduce the conversational training process closest to what OpenAI used?
Open on GitHub → Explain another repo

← nichtdax on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.