aether-models/qwen3-4b-instruct-2507

Qwen3 4B Instruct 2507

A Core AI bundle of Qwen/Qwen3-4B-Instruct-2507 for the Aether SDK (iOS and macOS 27+).

• Source: Qwen/Qwen3-4B-Instruct-2507 at revision cdbee75f17c01a7cc42f958dc650907174af0554, licence Apache-2.0. The licence is included as LICENSE.
• Changes from the source: converted from PyTorch to Core AI (.aimodel) by Aether forge (recipe qwen3-4b-instruct-2507@1). Weights are int8-linear-perchannel (8-bit weights). The tokenizer files are the source's own.

Variants

Variant	Platform	Arch	Compute	Compiled	Assets	Download
macos-any-gpu	macos	any	gpu	no (specialized on first load)	qwen3_4b_instruct_2507.aimodel 4.03 GB	4.04 GB
ios-h18p-gpu	ios	h18p	gpu	yes	qwen3_4b_instruct_2507.h18p.aimodelc 4.03 GB	4.04 GB

ios-h18p-gpu needs the com.apple.developer.kernel.increased-memory-limit entitlement.

Verification

Every row is a record in verification/ about exactly these bytes (matched by bundle digest). Reference rows are strict T2 passes of the unquantized export on the same fixture, in verification/reference/.

Variant	Tier	Result	Detail	Device	OS build	Compute	Record
ios-h18p-gpu	T0	pass		iPhone18,2	24A437	target	af78f59d
ios-h18p-gpu	T2	pass	17/18 strict; profile quantized-8bit; budgeted: think-prime; fixture 20fd9acbecb58090	iPhone18,2	24A437	target	c9b66240
macos-any-gpu	T0	pass		Mac17,6	26A428	target	ea689fe2
macos-any-gpu	T1	pass		Mac17,6	26A428	target	27108b24
macos-any-gpu	T2	pass	17/18 strict; 17/18 strict; profile quantized-8bit; budgeted: think-prime; fixture 20fd9acbecb58090	Mac17,6	26A428	target	e7df0192
unquantized reference (not published)	T2	pass	18/18 strict; 18/18 strict; profile strict; fixture 20fd9acbecb58090	Mac17,6	26A428	target	b8388bfc

The quantized profile also requires: T2 strict on the unquantized reference export (met by the reference row).

Use

aether run qwen3-4b-instruct-2507 --prompt "Hello"

import Aether

let aether = try Aether()
let chat = try await aether.chat("qwen3-4b-instruct-2507")
let reply = try await chat.respond(to: "Hello")
print(reply.text)