Language · Apple Silicon · Metal

Qwen3

Run the post-trained Qwen3-1.7B checkpoint in a compact native runtime. Qwen3-1.7B-Base is catalogued but not published.

Status: Available
Platform: Apple Silicon · Metal
Class: Language
Publisher: Alibaba (Qwen)
Architecture: Qwen3
Weights: Supplied by you

Guide

How to run it

Install
One signed package for Apple Silicon. Nothing else to install first.
step 1
```
$ ./install.sh
```
Activate this device
Approve the machine in your browser; it receives a licence bound to it.
step 2
```
$ veizik activate
```
Pull the checkpoint
Downloads the published safetensors into the local cache. No conversion.
step 3
```
$ veizik pull qwen3-1.7b
```

Run it

One-shot, interactive, or as a local HTTP endpoint.

step 4

$ veizik run qwen3-1.7b "The capital of France is"
$ veizik chat qwen3-1.7b
$ veizik serve qwen3-1.7b --port 8080

Support

Where it runs.

Evidence and availability are separate: a verified device is not a download, and an available model is not verified on every device.

VerifiedVerifiedRun end to end on this exact device, on a published build.Architecture-supportedArchitecture-supportedVerified on another device of the same architecture; this device itself is untested.Expected, untestedExpected, untestedThe engine targets this architecture; nobody has run this model on it. No claim is made.In preparationIn preparationNo build is downloadable yet, and no figure is offered.Not offeredNot offeredThis family is not offered on this platform.UnknownUnknownNo information yet — the platform is announced with the release.

Family	RTX 3080 Ti	RTX 3090	RTX 4090	RTX 5090	RTX A5000	A40	A100 80GB PCIe	A100-SXM4-80GB	H100 80GB HBM3	RTX PRO 4500 Blackwell	M4 Pro	M1 Ultra	M1 Max	M2 Pro
Qwen3 all variants	Not offered	Not offered	Not offered	Not offered	Not offered	Not offered	Not offered	Not offered	Not offered	Not offered	Verified	Verified	Verified	Verified

Availability by variant

Model id	Availability	Platform	Context	Weights in memory
`qwen3-1.7b`	Available	Apple Silicon	32,768	968.0 MB measured, deterministic load
`qwen3-1.7b-base`	Not published	Apple Silicon	—	—

Verified on

Apple M1 Max · build 2026.10.02-r10 · 2026-10-20 · published measurement
Apple M1 Ultra · build 2026.10.02-r10 · 2026-10-09 · published measurement
Apple M2 Pro · build 2026.10.02-r10 · 2026-10-06 · published measurement
Apple M4 Pro · build 2026.10.02-r10 · 2026-10-07 · published measurement

“Verified” means this family ran end to end on that exact device, on a published build. Architecture membership never upgrades a device to verified.

Variants

2 checkpoints

Model id	Size	Checkpoint	On disk	Context	Weights in memory
`qwen3-1.7b`	1.7B	Qwen/Qwen3-1.7B	3,281 MiB	32,768	968.0 MB
`qwen3-1.7b-base`	1.7B	Qwen/Qwen3-1.7B-Base	—	—	—

Performance

Measured on the released build

Model	Device	Decode tok/s	Prefill tok/s	Runs
Qwen3-1.7B	Apple M1 Ultra	289.8	4,213	11
Qwen3-1.7B	Apple M2 Pro	143.8	1,459	11
Qwen3-1.7B	Apple M1 Max	207.4	2,417	11
Qwen3-1.7B	Apple M4 Pro	204.3	2,161	11

512-token prompt, median of runs, lower bound. Every prompt length and device →

Licence

Weights and terms

Qwen3 weights are published by Alibaba under their own license. Veizik neither redistributes nor relicenses them.

Full licence terms →