Ollaya
Open-source daemon and CLI that pulls and serves open decision models locally behind a TypeSafe Jev-compatible API.
0 members list Ollaya, unchanged over the last 12 weeks, data as of 11 November 2026
Members who list Ollaya in their Stack, by week.
Members who list Ollaya
Be the first to add Ollaya
Nobody on Stackness runs Ollaya yet. Add it to your stack and you're the first member running it.
Homebrew analytics installs: 2 installs (2026) to 2 installs (2026)
Imported from Homebrew analytics, last reading Oct 2026
About Ollaya
Ollaya is an open-source daemon and command-line tool that downloads and runs decision models locally on a user's machine. It provides a TypeSafe-compatible API for serving various open decision models, including encoders and decoders, which can answer typed questions about text or JSON data with calibrated probability scores. The tool is designed for private inference, keeping data on the user's own hardware rather than sending it to external services.
Ollaya is developed by ollaya-dev and distributed under the Apache 2.0 license. Written in Rust, it supports a range of decision model families and provides features like language routing, configurable inference options, and batch question processing. The tool borrows its design philosophy from Ollama, offering a simple workflow to pull and run models behind a local API endpoint.
Checked 9 November 2026.
Nobody on Stackness runs Ollaya yet.
Sign up and add itPosts about Ollaya
- Strands Decider 2B: what it is and where it sits in an agent harness
Strands Decider 2B is AWS's free 1.9B decision model, out 1 October 2026: about 115 ms per question on an RTX 3090, consulted before each tool call.
- OpenAI Decisions API vs Jev: price per decision, latency, images and what switching costs
OpenAI Decisions costs 1.6 times Jev per decision and answers slower, but takes images without an experimental flag and comes with OpenAI's data terms.
- Mistral Large 4 vs Beam vs Kolibri: which open weights you can download, and what your hardware can serve
Only Kolibri's weights are out. Beam and Mistral Large 4 stay API previews until late October, and Kolibri's 4-bit build runs at 50 tok/s on a 64GB Mac.