M1 16GB

APPLE · 16GB unified memory · 3 reports

On the M1 16GB, Gemma 4 8B at 4-bit, the model most run on it, generates a median of 8.2 tokens per second from one plain run (community reports, as of 30 Oct 2026).

As of 30 Oct 2026, the models most run on the M1 16GB, with generation and prompt-processing medians from plain runs (one device, one request, no speculative decoding):

Gemma 4 8B
8.2gen t/s at 4-bit · one run
Qwen2.5 0.5B
95.6gen t/s at 4-bit · one run

Engines people use on it: Unsloth Studio 1 (reports) · llama.cpp 1 (reports)

See reported performance Estimate what fits

How to compare these reports →Filter reports by setup →All M1 Macs compared →

Run models on your M1 16GB? Add your numbers.

Your coding agent can run a speed test on your machine and send it here with a free account. Or paste a result you've already posted.

Benchmark with your agent Paste a result

Reported performance on the M1 16GB

These reports mix GPU counts and memory setups, so a model listed here may have needed extra cards or CPU offloading. The calculator estimates which models fit in 16 GB of unified memory.

Date ↓Date ↑Token speed ↓Token speed ↑Prefill ↓Prefill ↑Context ↓Context ↑Ranked

Qwen3.8 125B (6B active) Flash-Next

M1 16GB · Slotstream · 512 ctx

generation:: 1.8 tokens/s
quant:: 4-bit (MLX)
mtp (multi-token prediction):: off

Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is shown only when the source reports it; generation speed cannot tell us time to first token. Check the full setup before comparing.

User reports Qwen3.8 Flash-Next at 1.76 t/s generation on a Mac mini M1 16GB, streaming experts from the internal SSD. Setup is Slotstream 0.2.17 on MLX Swift with PipeNetwork's MLX 4-bit weights, one request, MTP off, a 512-token context and a 19-token prompt. The model's 105 GB sits on disk while the process peaks near 5 GB of memory. A second run gave the same speed.

Qwen2.5 0.5B

M1 16GB · llama.cpp

generation:: 95.6 tokens/s
quant:: Q4 (GGUF)

User benchmarks Qwen2.5-0.5B Q4 at 95.6 t/s decode with 18 ms TTFT on an Apple M1 16GB, averaged across defaults and normalized modes on the chat_turn workload. Setup is llama.cpp (llama-server) with Q4 GGUF weights, Metal backend, 1 warmup plus 4 measured reps per cell. LlamaStash leads raw llama-server (91.9 t/s / 20 ms), LM Studio (88.4 / 68) and Ollama 0.24.0 (79.6 / 102); in defaults mode LlamaStash reaches 99.0 t/s / 15 ms vs raw llama-server's 92.3 / 19, and normalized mode collapses the gap to under 1% (92.2 vs 91.5).

Gemma 4 8B E4B

M1 16GB · Unsloth Studio · 131,100 ctx

generation:: 8.2 tokens/s
quant:: 4bit (MLX)

codingagenticsummarization

User reports Gemma 4 E4B at around 8.2 tok/s on an M1 MacBook Air with 16GB unified memory. Setup is Unsloth Studio with a 4-bit MLX model and a 131.1k context window. The user says this speed is fine for chatbot use and asks for suggestions of 7B or 8B MLX models for coding and background tasks.

Get a weekly email of new M1 16GB reports and newly released models that fit it.

Email me new reports

Reported configurations

Individual observations from this page, not expected speeds or a ranking. Select a model to open its report, with the source and offloading conditions. Reported context may be a limit rather than actual input length.

Model / hardware	Quant / engine	Reported context	Generation
Qwen3.8 125B (6B active) Flash-Next M1 16GB	4-bit Slotstream	512	1.8 tokens/s
Qwen2.5 0.5B M1 16GB	Q4 llama.cpp	Not reported	95.6 tokens/s
Gemma 4 8B E4B M1 16GB	4bit Unsloth Studio	131,100	8.2 tokens/s

Frequently asked

How fast is the M1 16GB for local LLMs?

On the M1 16GB, Gemma 4 8B at 4-bit, the model most run on it, generates a median of 8.2 tokens per second from one plain run (community reports, as of 30 Oct 2026). Also reported: Qwen2.5 0.5B at 4-bit (95.6 t/s, one plain run).