Alternatives
Self-hosted alternatives to llama.cpp
Projects in the registry that can run on your own hardware and serve at least one of the same use cases as llama.cpp.
- Source
- Registry relations over fetched repository records — the evidence sits on each repository page
- Verified
- Evidence not verified
- Confidence
- High
01How this list was built
- Compared against
- llama.cpp
- Delivery
- Libraries
- Shared use cases
- Local LLM, Private LLM
- Selection rule
- open source, or self-hostable, and shares a use case
- Candidates found
- 16
02Alternatives
AnythingLLM
Hybrid — hosted or self-hosted
Desktop and server application that turns a document set into a chat workspace, with per-workspace embeddings, multiple model back ends and a built-in vector store.
- Repository
- Mintplex-Labs/anything-llm
- Stars
- 65,673
- Licence
- MIT
- In stacks
- 0
Shares: Local llm
Continue
Hybrid — hosted or self-hosted
Open-source IDE extension for VS Code and JetBrains that connects completion and chat to any model back end, including a local server, under a checked-in configuration file.
- Repository
- continuedev/continue
- Stars
- 35,622
- Licence
- Apache-2.0
- In stacks
- 1
Shares: Local llm
Jan
Hybrid — hosted or self-hosted
Open-source desktop assistant that runs open-weight models locally and can also call remote endpoints. The open alternative to LM Studio for a single-machine deployment.
- Repository
- janhq/jan
- Stars
- 44,559
- Licence
- Other
- In stacks
- 0
Shares: Local llm · Private llm
LibreChat
Self-hosted projects
Self-hosted multi-model chat application with authentication, per-conversation model switching, plugins, file upload and an admin configuration file. Familiar interface for staff moving off consumer tools.
- Repository
- danny-avila/LibreChat
- Stars
- 42,850
- Licence
- MIT
- In stacks
- 0
Shares: Private llm
LiteLLM
Cloud platforms
Gateway that presents one OpenAI-compatible API in front of many providers and local servers, with per-team keys, budgets, rate limits, fallbacks and request logging.
- Repository
- BerriAI/litellm
- Stars
- 57,216
- Licence
- Other
- In stacks
- 3
Shares: Private llm · Local llm
LMDeploy
Model servers
Serving toolkit from the InternLM team with its own inference engine and quantisation tooling, and strong coverage of Chinese open-weight model families.
- Repository
- InternLM/lmdeploy
- Stars
- 8,082
- Licence
- Apache-2.0
- In stacks
- 0
Shares: Local llm · Private llm
LM Studio
Hybrid — hosted or self-hosted
Desktop application for downloading and running open-weight models locally, with a chat interface and a local OpenAI-compatible server. Windows, macOS and Linux.
- Repository
- none linked
- Stars
- —
- Licence
- unknown
- In stacks
- 1
Shares: Local llm · Private llm
LocalAI
Model servers
Drop-in OpenAI-compatible API server that runs text, embedding, image and audio models locally across several back ends, including CPU-only deployments.
- Repository
- mudler/LocalAI
- Stars
- 49,166
- Licence
- MIT
- In stacks
- 0
Shares: Local llm · Private llm
NVIDIA NIM
Model servers
Prebuilt inference containers with an OpenAI-compatible API, deployable into your own cluster including air-gapped sites. Licensed through NVIDIA AI Enterprise rather than open source.
- Repository
- none linked
- Stars
- —
- Licence
- unknown
- In stacks
- 0
Shares: Local llm · Private llm
NVIDIA TensorRT-LLM
Libraries
Compiles models into optimised engines for NVIDIA GPUs. The build step is the trade: peak throughput on a fixed configuration, in exchange for recompiling when it changes.
- Repository
- nvidia/tensorrt-llm
- Stars
- 14,663
- Licence
- Other
- In stacks
- 0
Shares: Local llm · Private llm
Ollama
Model servers
Local model runtime with a one-command install, a model library and an OpenAI-compatible API. The usual starting point for running open-weight models on a workstation or small server.
- Repository
- ollama/ollama
- Stars
- 179,403
- Licence
- MIT
- In stacks
- 4
Shares: Local llm · Private llm
Open WebUI
Self-hosted projects
Self-hosted chat interface for local and hosted models, with user accounts, groups, document upload and built-in retrieval. Runs in Docker against Ollama, vLLM or any OpenAI-compatible endpoint.
- Repository
- open-webui/open-webui
- Stars
- 149,870
- Licence
- Other
- In stacks
- 5
Shares: Private llm · Local llm
Qdrant
Hybrid — hosted or self-hosted
Vector search engine written in Rust with payload filtering, hybrid search, quantisation and snapshots. Runs as a single container or a cluster, or as a managed cloud service.
- Repository
- qdrant/qdrant
- Stars
- 34,678
- Licence
- Apache-2.0
- In stacks
- 0
Shares: Private llm
SGLang
Model servers
Serving framework for large models with prefix caching and structured-output support, aimed at high-throughput deployments and multi-GPU serving.
- Repository
- sgl-project/sglang
- Stars
- 36,163
- Licence
- Apache-2.0
- In stacks
- 0
Shares: Local llm · Private llm
Tabby
Self-hosted projects
Self-hosted coding assistant with its own inference server, repository-aware context and IDE extensions. Runs entirely inside the network, including on a single GPU.
- Repository
- TabbyML/tabby
- Stars
- 33,834
- Licence
- Other
- In stacks
- 1
Shares: Local llm
vLLM
Model servers
High-throughput GPU inference server using paged attention and continuous batching, with an OpenAI-compatible API. Built for many concurrent users rather than single-session use.
- Repository
- vllm-project/vllm
- Stars
- 89,956
- Licence
- Apache-2.0
- In stacks
- 7
Shares: Local llm · Private llm
03Repository facts
| Repository | Stars | Health |
|---|---|---|
| Mintplex-Labs/anything-llmStop renting your intelligence. Own it with AnythingLLM. Everything you need for a powerful local-first agent experience | 66k | Maintenance 60%, Releases 60%, Contributors 50%, Issues 95%, Community 96%71 |
| continuedev/continueopen-source coding agent | 36k | Maintenance 60%, Releases 40%, Contributors 50%, Issues 74%, Community 91%64 |
| janhq/janJan is an open source alternative to ChatGPT that runs 100% offline on your computer. | 45k | Maintenance 80%, Releases 60%, Contributors 50%, Issues 88%, Community 93%76 |
| danny-avila/LibreChatEnhanced ChatGPT Clone: Features Agents, MCP, Skills, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini, Artifacts, AI model switching, message search, Code Interpreter, langchain, DALL-E-3, OpenAPI Actions, Functions, Secure Multi-User Auth, Presets, open-source for self-hosting. Active | 43k | Maintenance 60%, Releases 30%, Contributors 50%, Issues 83%, Community 93%63 |
| BerriAI/litellmThe fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM] | 57k | Maintenance 60%, Releases 60%, Contributors 50%, Issues 14%, Community 95%63 |
| InternLM/lmdeployLMDeploy is a toolkit for compressing, deploying, and serving LLMs. | 8.1k | Maintenance 80%, Releases 60%, Contributors 50%, Issues 26%, Community 78%66 |
| mudler/LocalAILocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required. | 49k | Maintenance 80%, Releases 80%, Contributors 50%, Issues 97%, Community 94%81 |
| nvidia/tensorrt-llmTensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in a performant way. | 15k | Maintenance 80%, Releases 40%, Contributors 50%, Issues 0%, Community 83%60 |
| ollama/ollamaGet up and running with Kimi-K2.6, GLM-5.2, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models. | 179k | Maintenance 60%, Releases 60%, Contributors 50%, Issues 79%, Community 100%70 |
| open-webui/open-webuiUser-friendly AI Interface (Supports Ollama, OpenAI API, ...) | 150k | Maintenance 60%, Releases 60%, Contributors 50%, Issues 99%, Community 100%72 |
| qdrant/qdrantQdrant - High-performance, massive-scale Vector Database and Vector Search Engine for the next generation of AI. Also available in the cloud https://cloud.qdrant.io/ | 35k | Maintenance 80%, Releases 60%, Contributors 50%, Issues 79%, Community 91%74 |
| sgl-project/sglangSGLang is a high-performance serving framework for large language models and multimodal models. | 36k | Maintenance 80%, Releases 80%, Contributors 50%, Issues 0%, Community 91%70 |
| TabbyML/tabbySelf-hosted AI coding assistant | 34k | Maintenance 40%, Releases 20%, Contributors 50%, Issues 90%, Community 91%55 |
| vllm-project/vllmA high-throughput and memory-efficient inference and serving engine for LLMs | 90k | Maintenance 60%, Releases 60%, Contributors 50%, Issues 22%, Community 99%65 |
04Ask about your own case
Which alternative is right depends on how many people will use it, what the data is and what you can operate. Ask and the answer is researched for your organisation.