7 open-source alternatives ranked by real activity

Open-source Groq alternatives

A curated, ranked list of the 7 best open-source alternatives to Groq.

The best open-source alternative to Groq is Ollama. If that doesn't suit you, other good options are vLLM, llama.cpp, SGLang and LocalAI.

Commercial alternatives to Groq include Cerebras, Fireworks AI, SambaNova and DeepInfra. See all 22.

Groq alternatives are mainly AI infrastructure tools, but some are also AI tools. 7 of them shipped code in the last 30 days, 7 can be self-hosted, and 7 use a permissive licence.

Last updated November 12, 2026 · ranked by GitHub stars, growth and recent commits

Ollama

A tool for downloading and running open large language models locally through a command line and REST API, with Docker support and many integrations.

GitHub stars
187k
Last commit
yesterday
Latest release
v0.40.16
Licence
MIT
Self-hosted
Yes
ollama.comOllama homepage screenshot

Ollama makes it straightforward to run open large language models on your own machine. You install it on macOS, Windows or Linux, or use the official Docker image, then pull a model and chat with it from the command line. It is written in Go and released under the MIT license, and it builds on the llama.cpp project for inference.

Besides the CLI, Ollama provides a REST API for managing and running models, with official Python and JavaScript client libraries. Models can be imported and customized through a Modelfile, and an online library lists what is available. It can be connected to coding agents and assistants such as Claude Code, OpenCode, Codex and OpenClaw, and many community chat interfaces, including Open WebUI and LibreChat, work with it.

Because models run on your own hardware, Ollama can serve as a self-hosted alternative to cloud assistants such as ChatGPT and Perplexity when privacy, offline use or cost control matter. It is aimed at developers and hobbyists experimenting with open models, and at teams that want to wire local models into their own applications.

Key features

  • Run open language models locally from the CLI
  • REST API for managing and running models
  • Official Python and JavaScript libraries
  • Modelfile for customizing and importing models
  • Official Docker image
  • Integrations with coding agents and chat interfaces

Pricing: Running models locally is free, and the Free plan includes starter cloud credits. Pro costs $20, Max $100 and Team $500 per month with included usage credits; Enterprise is custom.

vLLM

A Python library and server for fast, memory-efficient LLM inference and serving, with PagedAttention, continuous batching and broad quantization support.

GitHub stars
97k
Last commit
yesterday
Latest release
v0.33.0rc3
Licence
Apache-2.0
Self-hosted
Yes
vllm.aivLLM homepage screenshot

vLLM is an open-source library for running and serving large language models efficiently. It originated at UC Berkeley's Sky Computing Lab and is now developed by a large community of contributors from academia and industry. It is written in Python and released under the Apache-2.0 license.

Speed comes from techniques such as PagedAttention for attention key-value memory, batching incoming requests continuously, chunked prefill, prefix caching, CUDA and HIP graphs, and optimized attention and mixture-of-experts kernels. It supports many quantization formats, including FP8, INT8, INT4, GPTQ and AWQ, as well as speculative decoding and disaggregated prefill and decode.

On the usability side, vLLM integrates with Hugging Face models and supports parallel sampling, beam search, streaming outputs, structured outputs and tool calling, plus several parallelism modes (tensor, pipeline, data, expert and context) for distributed inference. It targets NVIDIA, AMD and TPU hardware and suits teams that serve open models in production and need high throughput.

Key features

  • PagedAttention and continuous batching
  • Quantization including FP8, INT8, GPTQ and AWQ
  • Speculative decoding support
  • Tensor, pipeline and expert parallelism
  • OpenAI-compatible API server
  • Hugging Face model integration
  • Streaming and structured outputs

Pricing: Free and open source under the Apache-2.0 license.

llama.cpp

A dependency-free C/C++ engine for running large language models locally on CPUs and GPUs, with quantization, a CLI and an HTTP server.

GitHub stars
134k
Last commit
yesterday
Latest release
v0.9.0
Licence
MIT
Self-hosted
Yes
llama.appllama.cpp homepage screenshot

llama.cpp is an open-source library and set of tools for running large language model inference, including vision-language models, with minimal setup. The aim is strong performance across many kinds of hardware, whether local or in the cloud. It is a plain C/C++ implementation without external dependencies, built on the ggml library, and is licensed under MIT.

The project optimizes for many platforms: Apple silicon through ARM NEON, Accelerate and Metal, x86 with AVX, AVX2, AVX512 and AMX, and several RISC-V extensions. Quantization from 1.5-bit to 8-bit integers reduces memory use and speeds up inference. GPU backends include CUDA for NVIDIA, HIP for AMD, MUSA, Vulkan and SYCL, and hybrid CPU plus GPU inference lets you partly accelerate models larger than available VRAM.

It provides a command-line interface and llama-server, which offers a REST API and a built-in web UI. You can install it from pre-built binaries, Docker or by building from source. Other local-AI tools such as Ollama build on it, which makes it relevant to developers embedding inference in their own applications and hobbyists running models on modest hardware.

Key features

  • Dependency-free C/C++ inference engine
  • Quantization from 1.5-bit to 8-bit
  • CUDA, HIP, Metal, Vulkan and SYCL backends
  • CPU plus GPU hybrid inference
  • llama-server with REST API and web UI
  • Optimized for Apple silicon and x86

Pricing: Free and open source under the MIT license.

SGLang

SGLang is an open-source framework for serving language, vision-language and diffusion models, built for fast inference at scale.

GitHub stars
39k
Last commit
yesterday
Latest release
v0.5.23
Licence
Apache-2.0
Self-hosted
Yes
sglang.ioSGLang homepage screenshot

SGLang is an inference framework, released as open source, for language models, vision-language models and diffusion models. It is optimized for agentic workloads, reinforcement-learning rollouts and large-scale serving. The project includes SGLang Diffusion, a built-in engine for image and video generation that ships in the same repository and Python package. It is released under the Apache-2.0 license.

Getting started takes either a prebuilt Docker image or a Python install using uv, followed by a launch command for the chosen model. A cookbook helps pick a model and hardware pair and generates a ready-to-run command. Supported hardware spans NVIDIA and AMD GPUs, Google TPUs, Intel GPUs and CPUs, Apple Silicon through Metal and MLX, and Huawei Ascend NPUs, with more integrations in progress. The wider SGLang ecosystem adds educational projects and community events.

Key features

  • Serving for LLMs, vision-language and diffusion models
  • Docker image and uv-based Python install
  • Cookbook with ready-to-run launch commands
  • Runs on NVIDIA, AMD, TPU, Intel and Ascend hardware
  • Built-in image and video generation engine
  • Tuned for agentic and RL rollout workloads

Pricing: Free and open source under the Apache-2.0 license.

LocalAI

An open-source AI engine that runs LLMs, vision, voice, image and video models on your own hardware behind OpenAI-compatible APIs, with no GPU required.

GitHub stars
50k
Last commit
yesterday
Latest release
v4.12.1
Licence
MIT
Self-hosted
Yes
localai.ioLocalAI homepage screenshot

LocalAI is an open-source AI engine for running models of many kinds, including language, vision, voice, image and video models, on your own hardware. A GPU is not required. It is written in Go and released under the MIT license, was started by Ettore Di Giacinto and is looked after by the LocalAI team.

Its design is a small core with separate backends that are pulled on demand, so nothing unused gets installed. Backends wrap engines such as llama.cpp, vLLM, whisper.cpp, stable-diffusion and MLX, and you can write your own in any language against an open interface. It can run on CPU alone, through Vulkan, or on NVIDIA, AMD, Intel and Apple Silicon hardware.

APIs are drop-in compatible with OpenAI, Anthropic and ElevenLabs across every backend. Multi-user features include API key authentication, user quotas and role-based access, and built-in agents support tool use, RAG, MCP and skills. The project stresses that data stays on your infrastructure. It suits developers and teams who want a private, self-hosted replacement for cloud AI APIs.

Key features

  • Runs LLM, vision, voice, image and video models
  • Composable backends pulled on demand
  • OpenAI, Anthropic and ElevenLabs API compatibility
  • Runs on CPU, NVIDIA, AMD, Intel and Apple Silicon
  • API key authentication, quotas and role-based access
  • Built-in agents with tools, RAG and MCP

Pricing: Free and open source under the MIT license.

Hugging Face Chat UI

SvelteKit chat interface for large language models that connects to any OpenAI-compatible API and powers HuggingChat.

GitHub stars
11k
Last commit
yesterday
Latest release
v0.10.0
Licence
Apache-2.0
Self-hosted
Yes
Hosted version
Available
huggingface.coHugging Face Chat UI homepage screenshot

Chat UI is a chat interface for LLMs built as a SvelteKit application. It is the open source codebase behind HuggingChat on Hugging Face and lets you run a similar conversational front end against models you choose, whether hosted or running locally.

The current version speaks only to OpenAI-compatible APIs, configured through a base URL and a models endpoint. The README lists the Hugging Face Inference Providers router, a llama.cpp server, Ollama through its OpenAI-compatible bridge, OpenRouter and Poe as example backends. Older provider-specific integrations were removed from the main branch but remain available on a legacy branch.

Setup involves creating an .env.local file with the endpoint and key, choosing a database option, and launching the app, with an optional Docker image. The project is written in TypeScript with Tailwind CSS and is licensed under Apache-2.0, making it a base for self-hosted assistants.

Key features

  • Chat front end for any OpenAI-compatible API
  • Works with llama.cpp, Ollama and OpenRouter
  • Hugging Face Inference Providers support
  • Configured through a simple .env.local file
  • Optional Docker image
  • SvelteKit and Tailwind CSS codebase

Pricing: Free and open source under the Apache-2.0 license.

Helicone

Helicone is an open-source LLM observability platform and AI gateway that logs requests, tracks cost and latency, and routes traffic across models.

GitHub stars
6.3k
Last commit
yesterday
Latest release
v2025.08.21-1
Licence
Apache-2.0
Self-hosted
Yes
Hosted version
Available
helicone.aiHelicone homepage screenshot

Helicone is an open-source platform for AI engineers that combines LLM observability with an AI gateway. It records the requests your application sends to model providers so you can trace agents, chatbots and document-processing pipelines, debug sessions, and see how much each call costs and how long it takes.

Integration is meant to be quick: a single line of code can log requests made through OpenAI, Anthropic, LangChain, Gemini, the Vercel AI SDK and similar tools. The gateway side offers access to many models through the OpenAI API format, with intelligent routing and automatic fallbacks. Other features listed in the README and repository topics include cost and latency tracking, datasets and fine-tuning support, prompt management, evaluation and a playground.

Helicone is licensed under Apache-2.0 and written in TypeScript. It can be self-hosted, and the vendor also runs a hosted cloud version. It suits teams that want visibility into LLM spend and quality without building their own logging pipeline.

Key features

  • One-line request logging for LLM providers
  • Agent and session tracing
  • Cost and latency tracking
  • AI gateway with routing and fallbacks
  • Prompt management and evaluation
  • Datasets for fine-tuning workflows

Pricing: Free Hobby plan with 10,000 requests. Pro costs $79 and Team $799 per month plus usage, each with a 7-day trial; Enterprise is custom.

Commercial Groq alternatives

Hosted, closed-source products that compete with Groq. Showing 12 of 22.

See every Groq alternative on its tool page

Groq alternatives: questions

What is the best open-source alternative to Groq?
Ollama is the top-ranked open-source alternative to Groq on Enlisted: A tool for downloading and running open large language models locally through a command line and REST API, with Docker support and many integrations. Other strong options are vLLM, llama.cpp, SGLang and LocalAI.
What are the best commercial alternatives to Groq?
Cerebras, Fireworks AI, SambaNova, DeepInfra and Together AI are among the hosted, closed-source alternatives to Groq listed on Enlisted, which lists 22 in total. Compare their pricing and features on each tool page.
Are these Groq alternatives free?
All 7 are open source, so the code is free to use under its licence, and all of them can be self-hosted on your own server or computer. 2 also offer a paid or managed cloud version if you'd rather not host it yourself.
How is this list of Groq alternatives ranked?
By a score built from GitHub stars, star growth over the last 30 days and how recently the code changed. 7 of these projects shipped code in the last 30 days. Data is refreshed daily, and nobody can pay to move up.

People also look for alternatives to…

View all