Pi coding agent with a local model: setup for Ollama, LM Studio or llama.cpp

Pi is a minimal, extensible open-source (MIT) coding agent for the terminal, with four core tools (read, write, edit, bash) and a very short system prompt. It has a built-in llama.cpp router integration and connects to Ollama, LM Studio and other OpenAI-compatible servers through ~/.pi/agent/models.json, which suits people who want a small prompt on a tight context budget.

Source code Official docs Compare all agents

At a glance

Type: cli
Licence: MIT
Latest release: 1.3.0 (2026-10-15)
Status: Maintained
Local backends: openai-compatible, anthropic-compatible, ollama, lmstudio, llamacpp
Needs native tool calling: Not verified
MCP support: Yes
Recommended minimum context: Not stated in its docs
Start-up prompt: ~600 tokens (single third-party measurement)
GitHub stars: 116.3k (2026-10-19)

How to set it up with a local model

Pick your backend. Each command has its own Copy button. Where a step uses an example model, swap in the one you run; the table below lists the ones that fit your machine.

Install it (macOS / Linux):

curl -fsSL https://pi.dev/install.sh | sh

Start Ollama with a 64k context window (or set the context slider in the Ollama app's settings).
```
OLLAMA_CONTEXT_LENGTH=64000 ollama serve
```
In a second terminal, pull a tool-capable coding model (replace qwen3-coder:30b with any model you prefer).
```
ollama pull qwen3-coder:30b
```
Quickest path: let Ollama install Pi if needed, configure it and start a session.
```
ollama launch pi
```

Or add Ollama to Pi's models file yourself (the dummy key makes the model available; Ollama ignores it).

~/.pi/agent/models.json

{
  "$schema": "https://pi.dev/schemas/models.schema.json",
  "providers": {
    "ollama": {
      "baseUrl": "http://localhost:11434/v1",
      "api": "openai-completions",
      "apiKey": "ollama",
      "models": [
        {
          "id": "qwen3-coder:30b"
        }
      ]
    }
  }
}

Make it the default provider and model.

~/.pi/agent/settings.json

{
  "defaultProvider": "ollama",
  "defaultModel": "qwen3-coder:30b"
}

Other ways to install

npm (Node.js 22.19+)

npm install -g --ignore-scripts @earendil-works/pi-coding-agent

Check that it works

Open /model inside Pi; your local model should be listed. For llama.cpp, check the router is reachable.

curl http://127.0.0.1:8080/health

Which local models fit, by machine

Checked by this site's engine at the context Pi coding agent needs, with 32 GB of system RAM for MoE offload. A Mac counts three quarters of unified memory as usable.

Machine	Model	Quant	Context	Fit	Ollama
12 GB GPU	Qwen 3.5 9B	Q4_K_M	32,768	Comfortable (7.4 GB of 12 GB)	`qwen3.5:9b`
	Qwen 3.6 35B-A3B	Q4_K_M	32,768	With CPU offload (MoE) (22.7 GB of 12 GB)	`qwen3.6:35b-a3b`
16 GB GPU	Qwen 3.5 9B	Q4_K_M	32,768	Comfortable (7.4 GB of 16 GB)	`qwen3.5:9b`
	Qwen 3.6 35B-A3B	Q4_K_M	32,768	With CPU offload (MoE) (22.7 GB of 16 GB)	`qwen3.6:35b-a3b`
24 GB GPU	Qwen3.8 27B	Q4_K_M	32,768	Comfortable (19.9 GB of 24 GB)	`qwen3.8:27b`
	Qwen 3.6 27B	Q4_K_M	32,768	Comfortable (19.9 GB of 24 GB)	`qwen3.6:27b`
	Devstral Small 2 24B	Q4_K_M	32,768	Comfortable (20.7 GB of 24 GB)	`devstral-small-2:24b`
	Qwen 3.6 35B-A3B	Q4_K_M	65,536	Tight (23.3 GB of 24 GB)	`qwen3.6:35b-a3b`
32 GB GPU	Qwen3.8 27B	Q4_K_M	32,768	Comfortable (19.9 GB of 32 GB)	`qwen3.8:27b`
	Qwen 3.6 27B	Q4_K_M	32,768	Comfortable (19.9 GB of 32 GB)	`qwen3.6:27b`
	Devstral Small 2 24B	Q4_K_M	32,768	Comfortable (20.7 GB of 32 GB)	`devstral-small-2:24b`
	Qwen 3.6 35B-A3B	Q4_K_M	65,536	Comfortable (23.3 GB of 32 GB)	`qwen3.6:35b-a3b`
	Qwen3-Coder 30B-A3B (MoE)	Q4_K_M	65,536	Comfortable (25.7 GB of 32 GB)	`qwen3-coder:30b`
Mac, 64 GB	Qwen 3.6 35B-A3B	Q8_0	131,072	Comfortable (40.7 GB of 48 GB)	—

Your exact machine: filter the matrix or check any model in the analyzer.

Common problems with local models

A models.json entry stays out of /model until Pi can resolve credentials for it, so keep a dummy apiKey for keyless local servers.
With --no-models-autoload, a llama.cpp model must be loaded through /llama before it appears in /model.
Pi's llama.cpp guide sets -c 32768; omitting -c uses the model's native context, which may need far more memory.
Pi does not set Ollama's context window; set it on the server with OLLAMA_CONTEXT_LENGTH or the app slider.

Other coding agents

How they rank for local models: the best agent harness for local LLMs.

Sources (7), checked 2026-10-20