Runtime comparison - verified
KoboldCpp vs llama.cpp
KoboldCpp (LostRuins (Concedo) / KoboldAI community) and llama.cpp (ggml-org), compared field by field from their official docs as of October 18, 2026. Their docs differ on 4 feature points, listed below; the table also shows versions, install methods and links.
Key differences
- License: KoboldCpp uses AGPL-3.0; llama.cpp uses MIT.
- Interface: KoboldCpp lists Desktop app and llama.cpp does not.
- Backends: llama.cpp lists SYCL, NPU and Other and KoboldCpp does not.
- Model formats: KoboldCpp lists Other and llama.cpp does not.
Generated from the fields below that differ. Versions, dates and links are left to the table.
Side by side
Every field we track, with the source linked on each value. Rows marked Differs are where the two runtimes' docs differ.
| Field | KoboldCpp | llama.cpp |
|---|---|---|
| VendorDiffers | LostRuins (Concedo) / KoboldAI community | ggml-org |
| LicenseDiffers | AGPL-3.0 | MIT |
| Open source | Yes | Yes |
| Latest releaseDiffers | v1.123.1 | v0.7.0 |
| Release dateDiffers | October 13, 2026 | October 15, 2026 |
| InterfaceDiffers | CLI, Desktop app, Server, Docker | CLI, Server, Docker |
| Platforms | Windows, Linux, macOS | macOS, Windows, Linux |
| BackendsDiffers | CPU, CUDA, Vulkan, Metal, ROCm | CPU, CUDA, ROCm, Metal, Vulkan, SYCL, NPU, Other |
| Model formatsDiffers | GGUF, Other | GGUF |
| OpenAI-compatible API | Yeshttp://localhost:5001/v1 (default port 5001, changed with --port) | Yesllama-server listens on http://127.0.0.1:9931 by default (--host 127.0.0.1, --port 9931); OpenAI-compatible routes are under /v1 |
| Anthropic-compatible API | Yes | Yes |
| Tool calling | Yes | Yes |
| Structured outputs | Yes | Yes |
| Vision (image input) | Yes | Yes |
| Multi-GPU | Yes | Yes |
| MCP client | Yes | Yes |
| Install methods listedDiffers | linux; windows; source | macOS / Linux (curl); Windows (PowerShell); macOS / Linux (Homebrew); Windows (Winget) |
| Run exampleDiffers | ./koboldcpp-linux-x64 --model /path/to/model.gguf | llama serve -hf ggml-org/Qwen3.5-0.8B-GGUF |
| Audience, per the vendorDiffers | People who want to run GGUF models locally from a single executable with no installation, using a bundled KoboldAI Lite web UI for chat, story writing and roleplay plus a local API for other apps. | Developers who want a dependency-free C/C++ inference engine, with a CLI and an OpenAI-compatible server, for running GGUF models on a wide range of CPUs and GPUs locally or in the cloud. |
| HomepageDiffers | github.com/LostRuins/koboldcpp | github.com/ggml-org/llama.cpp |
| Source repositoryDiffers | github.com/LostRuins/koboldcpp | github.com/ggml-org/llama.cpp |
Install both
Copied verbatim from the page linked under each command.
KoboldCpp
linux
curl -fLo koboldcpp-linux-x64 https://github.com/LostRuins/koboldcpp/releases/latest/download/koboldcpp-linux-x64 && chmod +x koboldcpp-linux-x64 && ./koboldcpp-linux-x64windows
cmd /c "curl -fLo koboldcpp.exe https://github.com/LostRuins/koboldcpp/releases/latest/download/koboldcpp.exe && koboldcpp.exe"source
git clone https://github.com/LostRuins/koboldcpp.git
cd koboldcppSource: github.com/LostRuins/koboldcpp#compiling-koboldcpp-from-source-code
llama.cpp
macOS / Linux (curl)
curl -LsSf https://llama.app/install.sh | shSource: github.com/ggml-org/llama.cpp
Windows (PowerShell)
irm https://llama.app/install.ps1 | iexSource: github.com/ggml-org/llama.cpp
macOS / Linux (Homebrew)
brew install llama.cppSource: github.com/ggml-org/llama.cpp/blob/master/docs/install.md
Windows (Winget)
winget install llama.cppSource: github.com/ggml-org/llama.cpp/blob/master/docs/install.md
FAQ
What is the difference between KoboldCpp and llama.cpp?
Going by their official docs: License: KoboldCpp uses AGPL-3.0; llama.cpp uses MIT. Interface: KoboldCpp lists Desktop app and llama.cpp does not. Backends: llama.cpp lists SYCL, NPU and Other and KoboldCpp does not. Model formats: KoboldCpp lists Other and llama.cpp does not.
Do KoboldCpp and llama.cpp both have an OpenAI-compatible API?
KoboldCpp: yes. http://localhost:5001/v1 (default port 5001, changed with --port). llama.cpp: yes. llama-server listens on http://127.0.0.1:9931 by default (--host 127.0.0.1, --port 9931); OpenAI-compatible routes are under /v1.
Which platforms do KoboldCpp and llama.cpp run on?
KoboldCpp: Windows, Linux, macOS. llama.cpp: macOS, Windows, Linux. Both list Windows, Linux and macOS.
Are KoboldCpp and llama.cpp open source?
KoboldCpp is open source (AGPL-3.0). llama.cpp is open source (MIT).
Do KoboldCpp and llama.cpp support tool calling and MCP?
Tool calling: KoboldCpp yes, llama.cpp yes. MCP client: KoboldCpp yes, llama.cpp yes. "Not stated" means the docs do not say.
What are the latest versions of KoboldCpp and llama.cpp?
As of October 18, 2026: KoboldCpp v1.123.1 (October 13, 2026) and llama.cpp v0.7.0 (October 15, 2026).
More comparisons
Newsletter
Local AI, every week
One email per week. New model releases, runtime updates and what they mean for what you build.
Keep exploring
Open-weight models you can run
Download sizes, licenses and the run command for each model.
All local LLM runtimes
Every runtime in one feature matrix, with sources.
Picking a runtime for coding agents
Ollama, LM Studio, vLLM and llama.cpp for agent workloads.
Setup guides
Point your coding tools at the model you picked.