Upstage builds large language models and document-processing engines for workplace workflows.
Mirai
Mirai Labs builds uzu, a Rust LLM engine for Apple devices, with compact Qwen, LFM, and Muse-Glimmer models.
About Mirai
Mirai Labs builds uzu, a Rust-based LLM engine for Apple Silicon, along with compact models for on-device use. The work covers iPhone, iPad, and Mac, with 4-bit and 8-bit versions of Qwen, LFM, and Muse-Glimmer models. The site describes benchmarks against MLX and llama.cpp.
You can use its models, runtime, and infrastructure to run AI locally on Apple devices. The stack includes local models, an inference engine, quantization, and an application layer focused on Apple device automation. The approach is designed for single-request, latency-sensitive execution with reduced memory use.
It suits developers and teams working in AI infrastructure who build Apple-based apps that need local inference. It is also relevant to those researching on-device execution, quantization, and model-runtime co-design for Apple Silicon.
Key features
- Rust LLM engine for iPhone, iPad and Mac
- 4-bit and 8-bit Qwen, LFM and Muse-Glimmer models
- Local inference designed for batch size = 1
- Quantization to reduce memory footprint
- Application layer for Apple device automation
- Benchmarks against MLX and llama.cpp
Best for: Developers building on-device AI features for iPhone, iPad, and Mac.
Good for
- Built specifically for Apple Silicon devices
- Combines models, engine and quantization in one stack
- Focuses on local execution on the device
Watch out for
- Limited to Apple devices
- Not open source
Mirai at a glance
- Website
- trymirai.com
- Type
- Online tool
- Pricing
- Not confirmed. Check the site
- Platforms
- iOS, Web
- Open source
- Not stated
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
Mirai alternatives
Tools and sites you could use instead, ranked by how closely they match.
Vercel’s AI SDK is a TypeScript toolkit for building AI applications and agents with React, Next.js, Vue, Svelte, and Node.js.
MLflow helps teams build AI products by debugging, evaluating, monitoring, and improving AI applications.
Open source
24 alternatives
Featherless is a serverless AI inference platform for open-weight models with API access to tens of thousands of models.
Paid
24 alternatives
LLM Gateway routes, manages, and analyzes requests across more than 40 language-model providers through one API.
LiteLLM provides a way to call more than 100 language models through an OpenAI-compatible interface.
Popular in AI Infrastructure & MLOps
Well-known listings from the same topic.
Guild AIguild.aiFlow AIflow-ai.comsystemprompt.iosystemprompt.ioCoreWeavecoreweave.comKili Technologykili-technology.comAPI docsdocs.x.ai
Mirai FAQ
What is uzu?
uzu is a Rust LLM engine by Mirai Labs for iPhone, iPad and Mac.
Which models does Mirai Labs provide?
It provides 4-bit and 8-bit Qwen, LFM and Muse-Glimmer models for on-device use.
Which devices does Mirai support?
The site focuses on Apple Silicon, including iPhone, iPad and Mac.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.