Skip to Content
Acknowledgements

Acknowledgements

Tiles utilizes the following open-source software.

We are thankful to the developers and maintainers of these libraries!

Library/ResourceDescriptionLicenseLink
ml-explore/mlx-lmRun large language models locally with MLX on Apple Silicon.MITmlx-lm
earendil-works/piAI agent toolkit with a unified LLM API, agent loop, TUI, and coding-agent CLI.MITpi
n0-computer/irohModular peer-to-peer networking stack in Rust for distributed apps.Apache-2.0iroh
sqlcipher/sqlcipherStandalone SQLite fork with AES-256 encryption and other database security features.BSD-3-ClauseSQLCipher
rusqlite/rusqliteErgonomic bindings to SQLite for Rust.MITrusqlite
ucan-wg/rs-ucanRust implementation of UCAN.Apache-2.0rs-ucan
atrium-rs/atrium (atrium-oauth)OAuth client library for AT Protocol.MITatrium-oauth
huggingface/hf-hubRust client for the Hugging Face Hub.Apache-2.0hf-hub
openai/harmonyRenderer for the Harmony response format used with gpt-oss.Apache-2.0harmony
openai-harmony (PyPI)Python package for OpenAI’s Harmony response format for gpt-oss.Apache-2.0openai-harmony
mozilla-ai/openresponses-pythonPython SDK for the OpenResponses specification.Apache-2.0openresponses-python
openresponses/openresponsesOpen specification for multi-provider, interoperable LLM interfaces.Apache-2.0OpenResponses
openai/gpt-oss-20bOpen-weight 20B language model from OpenAI.Apache-2.0gpt-oss-20b
mlx-community/gpt-oss-20b-MXFP4-Q4MLX-quantized gpt-oss-20b model for local inference.Apache-2.0gpt-oss-20b-MXFP4-Q4
openai/gpt-ossOpen-weight gpt-oss language models and reference tooling.Apache-2.0gpt-oss
lmstudio-ai/venvstacksVirtual environment stacks for Python.MITvenvstacks

This list is still a work in progress and will be updated as we continue to add more features and dependencies to Tiles.