Run private, real-time, and predictable inference workloads directly on users' devices.
Part 4: Brief history of Apple ML Stack
By Mirai Labs, frontier on-device AI lab. Building the models, inference runtime, and quantization stack from the device constraint up.
Mount Thor — AI Execution Environments on Apple Hardware
Managed macOS environments for AI workloads that require native desktop access, persistent state, or model inference on Apple silicon.
Optimizing On-Device Inference for Apple Silicon
A custom local engine that improves prefill and decode throughput
Introducing Pipette: A benchmarking suite for on-device intelligence — Blog
Meet Pipette, an open-source platform for reproducible on-device AI benchmarks across models, quantization, runtimes and hardware.
Part 3: iPhone Hardware and How It Powers On-Device AI
By Mirai Labs, frontier on-device AI lab. Building the models, inference runtime, and quantization stack from the device constraint up.
InferenceMAX™: Open Source Inference Benchmarking
NVIDIA GB200 NVL72, AMD MI355X, Throughput Token per GPU, Latency Tok/s/user, Perf per Dollar, Tokens per Provisioned Megawatt, DeepSeek R1 670B, GPTOSS 120B, Llama3 70B
Detail - Argmax
Detail, Apple's pick for iPad App of the Year 2025, leverages Argmax SDK to build their flagship AI features such as text-based video editing and automatic speaker switching using Argmax SDK, migrating from cloud APIs. - Dec 09, 2025
Open Source AI Inference Benchmark | InferenceX
Compare AI inference performance across GPUs and frameworks. Real benchmarks on NVIDIA GB200, B200, AMD MI355X, and more. Free, open-source, continuously updated.
Intelligence per Watt: Measuring Intelligence Efficiency of Local AI
Large language model (LLM) queries are predominantly processed by frontier models in centralized cloud infrastructure. Demand growth strains this paradigm faster than providers can scale. Two advances create an opportunity to rethink it: small, local LMs (<=20B active parameters) now achieve competitive performance to frontier models on many tasks, and local accelerators (e.g., Apple M4 Max) can host these models at interactive latencies. This raises the question: can local inference viably redistribute demand from centralized infrastructure? This requires measuring both whether local LMs can accurately answer real-world queries and whether they can do so efficiently on power-constrained devices (e.g., laptops). We propose intelligence per watt (IPW), task accuracy per unit of power, as a unified metric for the capability and efficiency of local inference across model-accelerator configurations. We evaluate 20+ state-of-the-art local LMs, 8 hardware accelerators (local and cloud), and 1M real-world single-turn chat and reasoning queries. For each query, we measure accuracy (local LM win rate against frontier models), energy, latency, and power. We find three key results. First, local LMs successfully answer 88.7% of these queries, with accuracy varying by domain. Second, longitudinal analysis from 2023-2025 shows IPW improved 5.3x, driven by both algorithmic and accelerator advances, with locally-serviceable query coverage rising from 23.2% to 71.3%. Third, local accelerators achieve at least 1.4x lower IPW than cloud accelerators running identical models, revealing significant headroom for local accelerator optimization. These findings demonstrate that local inference can meaningfully redistribute demand from centralized infrastructure for a substantial subset of queries, with IPW serving as the critical metric for tracking this transition.
Ivan Fioravanti ᯅ on Twitter / X
"We are releasing Open Source implementations for CoreAILanguageModel and MLXLanguageModel for running a myriad of local models on the Apple Neural Engine or your Mac's GPU" 👀 From #WWDC26: What’s new in the Foundation Models framework video: https://t.co/1NtKWYhNRs pic.twitter.com/HVtBsr3tjL— Ivan Fioravanti ᯅ (@ivanfioravanti) June 9, 2026
Introducing LFM2: The Fastest On-Device Foundation Models on the Market | Liquid AI
Today, we release LFM2, a new class of Liquid Foundation Models (LFMs) that sets a new standard in quality, speed, and memory efficiency for on-device deployment. Built on a hybrid architecture, LFM2 delivers 200% faster decode and prefill performance than Qwen3 and Gemma 3 on CPU. It also significantly outperforms models in each size class on instruction-following and function calling—the core capabilities that make LLMs reliable for building AI agents.
On-Device LLM Throughput Calculator - a Hugging Face Space by FL33TW00D-HF
This tool estimates and visualizes the throughput of Large Language Models on devices with memory bandwidth constraints. Users input device and model configurations, and the tool generates a plot s...
Darkbloom — Cost-Efficient Private AI Inference on Verified Macs
Encrypted inference on hardware-verified Apple Silicon. Comparable model performance, operator-blind privacy, and about 50% lower cost.
Darkbloom — Cost-Efficient Private AI Inference on Verified Macs
Encrypted inference on hardware-verified Apple Silicon. Comparable model performance, operator-blind privacy, and about 50% lower cost.