← Canonical rankings

M1 (16 GB) — benchmark record

M1 (16 GB) local LLM benchmarks on Apple Silicon Mac with 6 published rows across 6 models. Peak published speed is 58.0 tok/s on Phi-4 Mini Instruct 3.8B. Published runtimes include Ollama.

6Benchmark rows
6Models tested
58.0Fastest avg tok/s (Phi-4 Mini Instruct 3.8B)
0Lab benchmarks

Best published speed here is 58.0 tok/s on Phi-4 Mini Instruct 3.8B at Q4_K - Medium. This page is evidence, not the full buying answer.

Based on 6 external benchmarks; no lab runs yet.

Published runtimes: Ollama.

Published model coverage includes Phi-4 Mini Instruct 3.8B, Mistral 7B v0.3, Llama 3.1 8B, Qwen 3 8B plus 2 more published models. Published runtime coverage on this chip includes Ollama. Fastest published row is 58.0 tok/s on Phi-4 Mini Instruct 3.8B at Q4_K - Medium.

Other published M1 variants: 8 GB

This chip is part of a family. Compare all M1 RAM variants →

Raw benchmark rows for M1 (16 GB)

Rows sorted by avg tok/s descending. Click source badge to see original measurement page.

ModelQuantRAM req.ContextAvg tok/sPrompt tok/sRuntimeSource
Phi-4 Mini Instruct 3.8BQ4_K - Medium58.0 tok/sOllamaref
Mistral 7B v0.3Q4_K - Medium42.0 tok/sOllamaref
Llama 3.1 8BQ4_K - Medium40.0 tok/sOllamaref
Qwen 3 8BQ4_K - Medium38.0 tok/sOllamaref
DeepSeek R1 Distill Llama 8BQ4_K - Medium38.0 tok/sOllamaref
Qwen3.5-9BQ4_K - Medium35.0 tok/sOllamaref

Use Rankings when you need the best overall answer for a Mac, and keep this page open when you need the evidence behind that recommendation.

If you are comparing local Apple Silicon against rented cloud hardware, use AI Data Center Index for current GPU rental context.

benchmarks.json — full dataset  ·  chips.json — chip summaries  ·  benchmarks.csv — CSV export

Data from in-house lab measurements plus community-published benchmarks. See all chip families →