Best forPulling and running models with one command, with a local API that other tools already support.
Trade-offFewer tuning knobs than running the engine directly.
Used by 39 makers' products for this
For run models locally, the alternatives to gpt-oss that makers' products we track chose most are Ollama (39), Llama (21) and LM Studio (4). 8 of the 8 are open source or self-hostable (Ollama, Llama, LM Studio and 5 more).
Not a ranking: each option is described by the situation it fits, with its trade-off.
How many makers' products we track use each for run models locally.
| Option | Kind | Makers' products | Free tier | Open source | |
|---|---|---|---|---|---|
| Open-weight model | 0 | — | Yes | ||
| Local runtime (CLI) | 39 | — | Yes | ||
| Open-weight model | 21 | — | Self-hostable | ||
| Desktop app | 4 | — | Self-hostable | ||
| Inference engine | 1 | — | Yes | ||
| Desktop app | 0 | — | Yes | ||
| Inference engine | 0 | — | Yes | ||
| Open-weight model | 0 | — | Yes | ||
| Open-weight model | 0 | — | Yes |
App or engine. Desktop apps and CLIs download models for you and expose a local API in minutes. The underlying engine gives you full control over builds, quantization and flags, and you handle the details yourself.
gpt-oss (Open-weight model): OpenAI's open-weight reasoning models under Apache 2.0, with the 20B version running on a well-equipped laptop. Trade-off: Only two sizes, both text-only, and the 120B version needs a large GPU.
Best forPulling and running models with one command, with a local API that other tools already support.
Trade-offFewer tuning knobs than running the engine directly.
Used by 39 makers' products for this
Best forMeta's widely supported open-weight models, available in every local runtime and in many sizes.
Trade-offCustom license with use restrictions rather than an open-source license; no new Llama release since Llama 4 in April 2025, as Meta's developer platform now centres on its hosted Muse models.
Used by 21 makers' products for this
Best forBrowsing, downloading and chatting with models in a GUI, then serving them through a local API.
Trade-offFree but not open source.
Used by 4 makers' products for this
Best forFull control over builds, quantization and hardware flags, or embedding inference inside your own app.
Trade-offYou compile it, convert models and pick settings yourself.
Used by 1 maker's product for this
Best forAn open-source desktop app for downloading and chatting with local models, with a local OpenAI-compatible server.
Trade-offFewer model-management and tuning options than LM Studio.
No maker product tracked for this yet
Best forRunning and fine-tuning models on Apple silicon Macs, using unified memory efficiently.
Trade-offApple silicon only, and models need MLX-format weights.
No maker product tracked for this yet
Best forA broad open-weight family — chat, coding, vision and embeddings — in many sizes, mostly under Apache 2.0.
Trade-offDeveloped by Alibaba in China, which matters to some customers' procurement rules.
No maker product tracked for this yet
Best forGoogle's small and mid-size open models, built to run on laptops and phones, with Gemma 4 under Apache 2.0.
Trade-offEarlier Gemma versions carry Google's own terms of use, so check which version you ship.
No maker product tracked for this yet
It depends on what you need it for; we don't rank them. The one makers' products we track chose most for run models locally is Ollama (39 makers' products). You want a local model behind an API for your code and dev tools in minutes: Ollama. You'd rather browse and compare models in a desktop app before writing code: LM Studio and Jan. You ship local inference inside your own app or need settings the apps don't expose: llama.cpp.
None of the run models locally alternatives we list has a free tier.
Yes. For run models locally: Ollama, Llama (self-hostable), LM Studio (self-hostable), llama.cpp, Jan, MLX LM and 2 more.