Open Source BSD Large Language Models (LLM)
Sort By:
Large Language Models (LLM) for BSD
Browse free open source Large Language Models (LLM) and projects for BSD below. Use the toggles on the left to filter open source Large Language Models (LLM) by OS, license, language, programming language, and project status.
-
Grafana: The open and composable observability platformGrafana is the open source analytics & monitoring solution for every database.
-
Our Free Plans just got better! | Auth0You asked, we delivered! Auth0 is excited to expand our Free and Paid plans to include more options so you can focus on building, deploying, and scaling applications without having to worry about your security. Auth0 now, thank yourself later.
-
1
GLM-4.6
Agentic, Reasoning, and Coding (ARC) foundation models
GLM-4.6 is the latest iteration of Zhipu AI’s foundation model, delivering significant advancements over GLM-4.5. It introduces an extended 200K token context window, enabling more sophisticated long-context reasoning and agentic workflows. The model achieves superior coding performance, excelling in benchmarks and practical coding assistants such as Claude Code, Cline, Roo Code, and Kilo Code. Its reasoning capabilities have been strengthened, including improved tool usage during inference and more effective integration within agent frameworks. GLM-4.6 also enhances writing quality, producing outputs that better align with human preferences and role-playing scenarios. Benchmark evaluations demonstrate that it not only outperforms GLM-4.5 but also rivals leading global models such as DeepSeek-V3.1-Terminus and Claude Sonnet 4. -
2
llama.cpp
Port of Facebook's LLaMA model in C/C++
The llama.cpp project enables the inference of Meta's LLaMA model (and other models) in pure C/C++ without requiring a Python runtime. It is designed for efficient and fast model execution, offering easy integration for applications needing LLM-based capabilities. The repository focuses on providing a highly optimized and portable implementation for running large language models directly within C/C++ environments. -
3
GLM-4.5
GLM-4.5: Open-source LLM for intelligent agents by Z.ai
GLM-4.5 is a cutting-edge open-source large language model designed by Z.ai for intelligent agent applications. The flagship GLM-4.5 model has 355 billion total parameters with 32 billion active parameters, while the compact GLM-4.5-Air version offers 106 billion total parameters and 12 billion active parameters. Both models unify reasoning, coding, and intelligent agent capabilities, providing two modes: a thinking mode for complex reasoning and tool usage, and a non-thinking mode for immediate responses. They are released under the MIT license, allowing commercial use and secondary development. GLM-4.5 achieves strong performance on 12 industry-standard benchmarks, ranking 3rd overall, while GLM-4.5-Air balances competitive results with greater efficiency. The models support FP8 and BF16 precision, and can handle very large context windows of up to 128K tokens. Flexible inference is supported through frameworks like vLLM and SGLang with tool-call and reasoning parsers included. -
4
GPT4All
Run Local LLMs on Any Device. Open-source
GPT4All is an open-source project that allows users to run large language models (LLMs) locally on their desktops or laptops, eliminating the need for API calls or GPUs. The software provides a simple, user-friendly application that can be downloaded and run on various platforms, including Windows, macOS, and Ubuntu, without requiring specialized hardware. It integrates with the llama.cpp implementation and supports multiple LLMs, allowing users to interact with AI models privately. This project also supports Python integrations for easy automation and customization. GPT4All is ideal for individuals and businesses seeking private, offline access to powerful LLMs. -
Cloud-based help desk software with ServoDeskWhat if You Could Automate 90% of Your Repetitive Tasks in Under 30 Days? At ServoDesk, we help businesses like yours automate operations with AI, allowing you to cut service times in half and increase productivity by 25% - without hiring more staff.
-
5
DeepSeek-V3
Powerful AI language model (MoE) optimized for efficiency/performance
DeepSeek-V3 is a robust Mixture-of-Experts (MoE) language model developed by DeepSeek, featuring a total of 671 billion parameters, with 37 billion activated per token. It employs Multi-head Latent Attention (MLA) and the DeepSeekMoE architecture to enhance computational efficiency. The model introduces an auxiliary-loss-free load balancing strategy and a multi-token prediction training objective to boost performance. Trained on 14.8 trillion diverse, high-quality tokens, DeepSeek-V3 underwent supervised fine-tuning and reinforcement learning to fully realize its capabilities. Evaluations indicate that it outperforms other open-source models and rivals leading closed-source models, achieving this with a training duration of 55 days on 2,048 Nvidia H800 GPUs, costing approximately $5.58 million. -
6
DeepSeek R1
Open-source, high-performance AI model with advanced reasoning
DeepSeek-R1 is an open-source large language model developed by DeepSeek, designed to excel in complex reasoning tasks across domains such as mathematics, coding, and language. DeepSeek R1 offers unrestricted access for both commercial and academic use. The model employs a Mixture of Experts (MoE) architecture, comprising 671 billion total parameters with 37 billion active parameters per token, and supports a context length of up to 128,000 tokens. DeepSeek-R1's training regimen uniquely integrates large-scale reinforcement learning (RL) without relying on supervised fine-tuning, enabling the model to develop advanced reasoning capabilities. This approach has resulted in performance comparable to leading models like OpenAI's o1, while maintaining cost-efficiency. To further support the research community, DeepSeek has released distilled versions of the model based on architectures such as LLaMA and Qwen. -
7
ChatGLM-6B
ChatGLM-6B: An Open Bilingual Dialogue Language Model
ChatGLM-6B is an open bilingual (Chinese + English) conversational language model based on the GLM architecture, with approximately 6.2 billion parameters. The project provides inference code, demos (command line, web, API), quantization support for lower memory deployment, and tools for finetuning (e.g., via P-Tuning v2). It is optimized for dialogue and question answering with a balance between performance and deployability in consumer hardware settings. Support for quantized inference (INT4, INT8) to reduce GPU memory requirements. Automatic mode switching between precision/memory tradeoffs (full/quantized). -
8
DeepSeek LLM
DeepSeek LLM: Let there be answers
The DeepSeek-LLM repository hosts the code, model files, evaluations, and documentation for DeepSeek’s LLM series (notably the 67B Chat variant). Its tagline is “Let there be answers.” The repo includes an “evaluation” folder (with results like math benchmark scores) and code artifacts (e.g. pre-commit config) that support model development and deployment. According to the evaluation files, DeepSeek LLM 67B Chat achieves strong performance on math benchmarks under both chain-of-thought (CoT) and tool-assisted reasoning modes. The model is trained from scratch, reportedly on a vast multilingual + code + reasoning dataset, and competes with other open or open-weight models. The architecture mirrors established decoder-only transformer families: pre-norm structure, rotational embeddings (RoPE), grouped query attention (GQA), and mixing in languages and tasks. It supports both “Base” (foundation model) and “Chat” (instruction / conversation tuned) variants. -
9
VibeThinker
Diversity-driven optimization and large-model reasoning ability
VibeThinker is a compact but high-capability open-source language model released by WeiboAI (Sina AI Lab). It contains about 1.5 billion parameters, far smaller than many “frontier” models, yet it is explicitly optimized for reasoning, mathematics, and code generation tasks rather than general open-domain chat. The innovation lies in its training methodology: the team uses what they call the Spectrum-to-Signal Principle (SSP), where a first stage emphasizes diversity of reasoning paths (the “spectrum” phase) and a second stage uses reinforcement techniques (the “signal” phase) to refine toward correctness and strong reasoning. The result is a model that outpaces many much larger models on domain-specific benchmarks, demonstrating that smaller models, if trained carefully and with the right objectives, can achieve high performance in reasoning-centric tasks. -
Free and Open Source HR SoftwareGive your HR team the tools they need to streamline administrative tasks, support employees, and make informed decisions with the OrangeHRM free and open source HR software.
-
10
Dramatron
Dramatron uses large language models to generate coherent scripts
Dramatron is an interactive co-writing tool developed by Google DeepMind that leverages large language models to help authors create screenplays and theatre scripts. It uses a hierarchical story generation approach to maintain coherence and structure across multiple levels of a narrative, from a single logline to detailed character descriptions, locations, plot points, and dialogue. Dramatron operates as a creative assistant rather than a fully autonomous system, offering human writers material to edit, adapt, and reinterpret. It was evaluated through user studies with professional playwrights and screenwriters, who found it particularly valuable for world-building, idea generation, and exploring alternative plotlines. The system can be run locally or in Google Colab, where users can integrate their own large language models by implementing sampling functions. -
11
LLM Datasets
Curated list of datasets and tools for post-training
LLM Datasets curates and standardizes datasets commonly used to train and fine-tune large language models, reducing the overhead of hunting down sources and normalizing formats. The repository aims to make datasets easy to inspect and transform, with scripts for downloading, deduping, cleaning, and converting to formats like JSONL that slot into training pipelines. It highlights instruction-tuning and conversation-style corpora while also pointing to code, math, or domain-specific sets for targeted capabilities. Quality is a recurring theme: examples and utilities help filter low-value samples, enforce length limits, and split train/validation consistently so results are comparable. Licensing and provenance are surfaced to encourage compliant usage and to guide dataset selection in commercial settings. For practitioners, the repo is a practical “starting pantry” that accelerates experimentation and helps keep data wrangling from dominating the project timeline. -
12
files-to-prompt
Concatenate a directory full of files into a single prompt
files-to-prompt is a Python command-line tool that takes one or more files or entire directories and concatenates their contents into a single, LLM-friendly prompt. It walks the directory tree, outputting each file preceded by its relative path and a separator, so a model can understand which content came from where. The tool is aimed at workflows where you want to ask an LLM questions about a whole codebase, documentation set, or notes folder without manually copying files together. It includes rich filtering controls, letting you limit by extension, include or skip hidden files, and ignore paths that match glob patterns or .gitignore rules. The output format is flexible: you can emit plain text, Markdown with fenced code blocks, or a Claude-XML style format designed for structured multi-file prompts. It can read file paths from stdin (including NUL-separated paths), which makes it easy to combine with find, rg, or other shell tools. -
13
GLM-130B
GLM-130B: An Open Bilingual Pre-Trained Model (ICLR 2023)
GLM-130B is an open bilingual (English and Chinese) dense language model with 130 billion parameters, released by the Tsinghua KEG Lab and collaborators as part of the General Language Model (GLM) series. It is designed for large-scale inference and supports both left-to-right generation and blank filling, making it versatile across NLP tasks. Trained on over 400 billion tokens (200B English, 200B Chinese), it achieves performance surpassing GPT-3 175B, OPT-175B, and BLOOM-176B on multiple benchmarks, while also showing significant improvements on Chinese datasets compared to other large models. The model supports efficient inference via INT8 and INT4 quantization, reducing hardware requirements from 8× A100 GPUs to as little as a single server with 4× RTX 3090s. Built on the SwissArmyTransformer (SAT) framework and compatible with DeepSpeed and FasterTransformer, it supports high-speed inference (up to 2.5× faster) and reproducible evaluation across 30+ benchmark tasks. -
14
LangChain
⚡ Building applications with LLMs through composability ⚡
Large language models (LLMs) are emerging as a transformative technology, enabling developers to build applications that they previously could not. But using these LLMs in isolation is often not enough to create a truly powerful app - the real power comes when you can combine them with other sources of computation or knowledge. This library is aimed at assisting in the development of those types of applications. -
15
MiniMax-M1
Open-weight, large-scale hybrid-attention reasoning model
MiniMax-M1 is presented as the world’s first open-weight, large-scale hybrid-attention reasoning model, designed to push the frontier of long-context, tool-using, and deeply “thinking” language models. It is built on the MiniMax-Text-01 foundation and keeps the same massive parameter budget, but reworks the attention and training setup for better reasoning and test-time compute scaling. Architecturally, it combines Mixture-of-Experts layers with lightning attention, enabling the model to support a native context length of 1 million tokens while using far fewer FLOPs than comparable reasoning models for very long generations. The team emphasizes efficient scaling of test-time compute: at 100K-token generation lengths, M1 reportedly uses only about 25 percent of the FLOPs of some competing models, making extended “think step” traces more feasible. M1 is further trained with large-scale reinforcement learning over diverse tasks. -
16
Tencent-Hunyuan-Large
Open-source large language model family from Tencent Hunyuan
Tencent-Hunyuan-Large is the flagship open-source large language model family from Tencent Hunyuan, offering both pre-trained and instruct (fine-tuned) variants. It is designed with long-context capabilities, quantization support, and high performance on benchmarks across general reasoning, mathematics, language understanding, and Chinese / multilingual tasks. It aims to provide competitive capability with efficient deployment and inference. FP8 quantization support to reduce memory usage (~50%) while maintaining precision. High benchmarking performance on tasks like MMLU, MATH, CMMLU, C-Eval, etc. -
17
GPT-NeoX
Implementation of model parallel autoregressive transformers on GPUs
This repository records EleutherAI's library for training large-scale language models on GPUs. Our current framework is based on NVIDIA's Megatron Language Model and has been augmented with techniques from DeepSpeed as well as some novel optimizations. We aim to make this repo a centralized and accessible place to gather techniques for training large-scale autoregressive language models, and accelerate research into large-scale training. For those looking for a TPU-centric codebase, we recommend Mesh Transformer JAX. If you are not looking to train models with billions of parameters from scratch, this is likely the wrong library to use. For generic inference needs, we recommend you use the Hugging Face transformers library instead which supports GPT-NeoX models. -
18
GenAI Agents
Implementations for various Generative AI Agent techniques
GenAI Agents is a large, tutorial-driven repository that teaches you how to design, build, and experiment with generative AI agents. It spans a spectrum from simple conversational bots and basic question-answering agents to complex multi-agent systems that coordinate on research, education, business workflows, and creative tasks. The implementations leverage modern frameworks such as LangChain, LangGraph, AutoGen, PydanticAI, CrewAI, and more, showing how each can be wired into realistic agent workflows. The repo is structured by categories like beginner agents, framework tutorials, educational agents, business agents, creative agents, analysis agents, news bots, shopping assistants, task management agents, QA bots, and advanced systems such as controllable RAG agents. For each agent, you typically get an overview, implementation notes, and external resources (blog posts, videos, documentation) to deepen understanding. -
19
Purple Llama
Set of tools to assess and improve LLM security
Purple Llama is an umbrella safety initiative that aggregates tools, benchmarks, and mitigations to help developers build responsibly with open generative AI. Its scope spans input and output safeguards, cybersecurity-focused evaluations, and reference shields that can be inserted at inference time. The project evolves as a hub for safety research artifacts like Llama Guard and Code Shield, along with dataset specs and how-to guides for integrating checks into applications. CyberSecEval, one of its flagship components, provides repeatable evaluations for security risk, including agent-oriented tasks such as automated patching benchmarks. The aim is to make safety practical: ship testable baselines, publish metrics, and provide drop-in implementations that reduce friction for teams adopting Llama. Documentation and sites attached to the repo walk through setup, usage, and the rationale behind each safeguard, encouraging community contributions. -
20
Qwen2.5
Open source large language model by Alibaba
Qwen2.5 is a series of large language models developed by the Qwen team at Alibaba Cloud, designed to enhance natural language understanding and generation across multiple languages. The models are available in various sizes, including 0.5B, 1.5B, 3B, 7B, 14B, 32B, and 72B parameters, catering to diverse computational requirements. Trained on a comprehensive dataset of up to 18 trillion tokens, Qwen2.5 models exhibit significant improvements in instruction following, long-text generation (exceeding 8,000 tokens), and structured data comprehension, such as tables and JSON formats. They support context lengths up to 128,000 tokens and offer multilingual capabilities in over 29 languages, including Chinese, English, French, Spanish, and more. The models are open-source under the Apache 2.0 license, with resources and documentation available on platforms like Hugging Face and ModelScope. This is a full ZIP snapshot of the Qwen2.5 code. -
21
CodeLlama
Inference code for CodeLlama models
Code Llama is a family of Llama-based code models optimized for programming tasks such as code generation, completion, and repair, with variants specialized for base coding, Python, and instruction following. The repo documents the sizes and capabilities (e.g., 7B, 13B, 34B) and highlights features like infilling and large input context to support real IDE workflows. It targets both general software synthesis and language-specific productivity, offering strong performance among open models at release time. Typical usage includes prompt-driven generation, function or class completion, and zero-shot adherence to natural-language instructions about code changes. The ecosystem provides multiple distributions (e.g., HF format) so developers can integrate with standard toolchains and serving stacks. As part of the broader Llama effort, Code Llama complements instruction-tuned chat models by focusing on code-centric tasks and editor integrations. -
22
FastEdit
Editing large language models within 10 seconds
FastEdit focuses on rapid “model editing,” letting you surgically update facts or behaviors in an LLM without full fine-tuning. It implements practical editing algorithms that insert or revise knowledge with targeted parameter updates, aiming to preserve model quality outside the edited scope. This approach is valuable when you need urgent corrections—think product names, APIs, or fast-changing facts—without retraining on large corpora. The repository provides evaluation harnesses so you can measure locality (does the change stay contained?) and generalization (does the change apply where it should?). It’s structured for repeatable experiments, making side-by-side comparisons of editing methods and hyperparameters straightforward. For applied teams, FastEdit offers a toolbox to keep models current and compliant while minimizing collateral damage to overall performance. -
23
LLM CLI
Access large language models from the command-line
A CLI utility and Python library for interacting with Large Language Models, both via remote APIs and models that can be installed and run on your own machine. -
24
LLM Course
Course to get into Large Language Models (LLMs)
LLM Course is a hands-on, notebook-driven path for learning how large language models work in practice, from data curation to training, fine-tuning, evaluating, and deploying. It emphasizes reproducible experiments: each step is demonstrated with runnable code, clear dependencies, and references to commonly used open-source models and libraries. Learners get exposure to multiple adaptation strategies—LoRA/QLoRA, instruction fine-tuning, and alignment techniques—so they can choose approaches that fit their hardware and budgets. The materials also cover inference optimization and quantization to make serving LLMs feasible on commodity GPUs or even CPUs, which is crucial for side projects and startups. Evaluation is treated as a first-class topic, with examples of automatic and human-in-the-loop methods to catch regressions and verify quality beyond simple loss values. By the end, students have a mental model and a practical toolkit for iterating on datasets, training configs, etc. -
25
LlamaIndex
Central interface to connect your LLM's with external data
LlamaIndex (GPT Index) is a project that provides a central interface to connect your LLM's with external data. LlamaIndex is a simple, flexible interface between your external data and LLMs. It provides the following tools in an easy-to-use fashion. Provides indices over your unstructured and structured data for use with LLM's. These indices help to abstract away common boilerplate and pain points for in-context learning. Dealing with prompt limitations (e.g. 4096 tokens for Davinci) when the context is too big. Offers you a comprehensive toolset, trading off cost and performance.
