Skip to main content

LemonPie

PyPI Version GitHub License Python Version from PEP 621 TOML

LemonPie — A lightweight CLI for LLM inference servers

LemonPie is a compact CLI for chatting with any OpenAI-compatible inference server — Ollama, llama.cpp, LM Studio, vLLM, and others. It manages chat sessions and persists history locally, designed to be tiny and portable — ideal for lightweight systems and terminal users, but usable anywhere with Python and network access.

✨ Features

  • Works with any OpenAI-compatible backend — Ollama, llama.cpp, LM Studio, vLLM, and more
  • Start and manage chat sessions with any configured model
  • Persist session history in your OS's standard user data directory
  • Configure host, backend, API key, default model, and aliases via a config file in your OS's standard user config directory
  • Interactive setup wizard (lmConfig --init) — select backend, configure server, verify reachability, and register models
  • Add, remove, and list models with aliases (lmConfig)
  • Switch models mid‑session with confirmation
  • Atomic writes for safety (tmp + replace)
  • Streaming responses for real‑time output
  • Graceful handling of Ctrl‑C interrupts

📂 Project Structure

LemonPie/
├── src/
│   └── lemonpie/            # Core package namespace
│       ├── main.py          # CLI session manager (entrypoints: lm, lemonpie)
│       ├── config.py        # Configuration logic & I/O
│       ├── paths.py         # OS-appropriate config/data directory resolution
│       ├── time_utils.py    # Timestamp helpers
│       ├── client.py        # HTTP client & error handling
│       ├── models.py        # Model finding & resolution
│       ├── cli/
│       │   ├── parser.py    # Argument parsing (lm)
│       │   ├── ui.py        # Terminal helpers & confirmation
│       │   └── config_cmd.py # Config manager (entrypoint: lmConfig)
│       └── sessions/
│           └── storage.py   # Session persistence
├── tests/                   # pytest suite
├── docs/
│   ├── config.example.json  # Configuration template
│   └── design/               # Design docs (model-verification.md, etc.)
├── .github/workflows/       # CI & release pipelines
├── pyproject.toml           # Package & build configuration
├── README.md                # Project overview
├── LICENSE                  # GPLv3 license
└── .gitignore

⚙️ Configuration

LemonPie keeps its config file and session history in your OS's standard user directories (via platformdirs) — not in the project folder:

OS Config Path Sessions Path
Linux ~/.config/lemonpie/ ~/.local/share/lemonpie/sessions/
macOS ~/Library/Application Support/lemonpie/ ~/Library/Application Support/lemonpie/sessions/
Windows %LOCALAPPDATA%\lemonpie\ %LOCALAPPDATA%\lemonpie\sessions\

All three are overridable via environment variables — LEMONPIE_HOST, LEMONPIE_CONFIG_DIR, and LEMONPIE_SESSION_DIR — useful for testing or running in a container.

docs/config.example.json is a template. Run lmConfig --init to set up your config interactively — it verifies server reachability, lists installed models for selection, and writes the file for you. lmConfig also auto-creates a minimal config on first use if you prefer to configure manually.

Example structure:

{
  "backend": "ollama",
  "server": "http://localhost:11434",
  "api_key": null,
  "default": "qwen2.5-coder:3b",
  "models": [
    { "alias": "qwen", "name": "qwen2.5-coder:3b" },
    { "alias": null,   "name": "qwen2.5-coder:1.5b-base" }
  ]
}

backend: which inference server you're using (e.g. ollama, llamacpp, lmstudio)
server: base URL of your inference server
api_key: optional API key for hosted backends (e.g. OpenAI)
default: canonical model name (full identifier)
models: list of objects with alias (nullable) and name

🚀 Usage

lm / lemonpie

  • Start a new session
    lm -n "Hello"

  • Start with a specific model
    lm -n -m qwen "Hello"

  • Send a prompt to the current session
    lm "Tell me about git add"

  • List sessions
    lm -l

  • Print current session history
    lm -p

  • Show what would happen without sending requests or modifying any data (dry run)
    lm --dry-run -n "Hello"

  • Switch model in current session
    lm -m qwen2.5-coder:3b

lmConfig

  • First-time setup
    lmConfig --init

  • First-time setup (skip server verification)
    lmConfig --init --force

  • Manage models
    lmConfig --add-model qwen qwen2.5-coder:3b
    lmConfig --list-models
    lmConfig --remove-model qwen

👉 Full command reference.

🗺 Roadmap

Planned features and improvements:

  • Code completion (autocomplete)
  • Model verification state machine
  • Cross‑platform packaging (apt, winget, Homebrew)
  • Future schema validation for config/session files

👉 See the full roadmap for details.

📜 License

This project is licensed under the GNU General Public License v3.0 (GPLv3).You may redistribute and/or modify it under the terms of the GPL as published by the Free Software Foundation.See the license file for details.

Metadata

Release files for lemonpie-cli 0.3.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for lemonpie-cli 0.3.1
File Size Uploaded Provenance Actions
lemonpie_cli-0.3.1.tar.gz 52.9 kB With provenance Details for lemonpie_cli-0.3.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for lemonpie-cli 0.3.1
File Interpreter ABI Platform Actions
lemonpie_cli-0.3.1-py3-none-any.whl Python 3 none any Details for lemonpie_cli-0.3.1-py3-none-any.whl

Total release size: 85.9 kB

Release files / lemonpie_cli-0.3.1.tar.gz

Download URL lemonpie_cli-0.3.1.tar.gz
Size 52.9 kB
Tags Source
SHA-256 checksum
How to use checksums
a613e7475d9d5f9913cefbaf4f49897ada6f9a4c7af2c276cb241e0c38d468f5
BLAKE2b-256 checksum
How to use checksums
93b44ccf12c5c7bb95b87e118c8ef6ba8247ed8799d1775a2bb714c474142459
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 15, 2026.

Transparency log

Release files / lemonpie_cli-0.3.1-py3-none-any.whl

Download URL lemonpie_cli-0.3.1-py3-none-any.whl
Size 33.1 kB
Tags Python 3
SHA-256 checksum
How to use checksums
ae2c21038d56a9b0c45a00baeeeb163304191a47ca0417d63168422843c4e832
BLAKE2b-256 checksum
How to use checksums
de50a09aa8a925bc94bd6f31724b3fad782e47f17d75a5f9a9ae3a16e0f0235d
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 15, 2026.

Transparency log

Security for lemonpie-cli 0.3.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

Both files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for lemonpie_cli-0.3.1.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 15, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page