Arm-optimized Qwen3.5-2B has 1.882 billion total parameters. The selected Q4_K_M GGUF file is 1.274 GB.
Data snapshot: . 1 published test in this catalog.
Model file and details
Provider
Arm / Qwen
Total parameters
1,881,825,088
Exact file size
1,274,396,288 bytes
Quantization / precision
Q4_K_M
Weight repository
Arm/qwen3-5-2b-q4-k-m-ggml-llama-cpp-vivo-x300
Original model repository
Qwen/Qwen3.5-2B
Weight revision
ef3036fe7fd026eb7b39493d950948d3c1caf19e
Model revision
15852e8c16360a2fea060d615a32b45270f8a8fc
Runtime notes
Confirm model architecture and quantization support in your runner.
Recorded release date
2026-10-13
Date evidence
Arm published this Apache-2.0, llama.cpp Q4_K_M release on 13 October 2026 for the official Qwen/Qwen3.5-2B source checkpoint. The pinned repository includes the exact language-model artifact, an Arm optimization recipe and artifact-specific measurements on one vivo X300; the separate vision projector is unquantized and was not benchmarked.
These scores were compiled from the linked sources. These published scores were not independently tested here. Different prompts, precision, thinking budgets and tools can change results.
MMLU-Redux
62.458 / 100
Revised MMLU knowledge test. Kept separate from original MMLU and MMLU-Pro.
Benchmark efficiency: 49.01 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Arm publisher evaluation of this exact Q4_K_M language-model artifact on the edinburgh-dawg/mmlu-redux-2.0 test split: 5,330 questions, 0-shot, single-token letter-logit scoring in llama.cpp. Arm measured a 63.9024% F16 baseline and 62.4578% for Q4_K_M, a -1.4446 percentage-point change.
Publisher measurement, not independent validation. The same release reports 22.085 output tokens/s, 983.288 ms time to first token, 6,782.289 ms median end-to-end latency and 2,523.973 MB peak memory on one vivo X300 CPU with four threads, batch size 1, 128 prompt tokens, 128 generated tokens, five warm-up runs and 20 measured runs. Tests waited for Android thermal status 0 and used fixed performance mode. Those results are specific to that device, build and workload; they do not establish speed or compatibility on other Arm systems. The 1,274,396,288-byte file contains the text decoder only. Vision requires a separate 668,226,592-byte F16 projector that Arm says was not quantized, benchmarked or evaluated. Weight size excludes runtime and context-cache headroom.
Context and version evidence: The repository and file name identify the selected model. Revision hashes are shown above when recorded. Tested context length, full hardware configuration and run date are not consistently recorded in this snapshot. Only settings explicitly stated in each benchmark record are known.
The 1.274 GB figure covers model weights in this file. It excludes the context cache, runtime, operating system and any separate multimodal files. It is not a measured RAM requirement or a token-generation speed.
These complete configurations pass the initial weight-size check for the selected 1.27 GB file. Allow extra memory for context and software. Check the model’s runtime notes and your inference software before buying.
Listing evidence: 2026-09-25. Amazon.com base price visible in an owner-supplied screenshot reviewed 25 Sep 2026; capture time not supplied. Dated reference, not a live quote. Excludes card signup offers, coupons, tax and shipping.
Listing evidence: 2026-10-03. Apple US launch price announced 22 September 2026: USD 1,699. Dated manufacturer reference, not an Amazon offer; tax and shipping may differ.
Listing evidence: 2026-09-24. GMKtec, EVO-X2 / 64GB / 1TB, recorded 22 Sep 2026; exact ASIN capacity identified 24 Sep. This is a dated reference, not a current Amazon price; tax, shipping and bundle contents may differ.
Memory matches are estimates, not tested compatibility or speed. As an Amazon Associate I earn from qualifying purchases.
Confirm model architecture and quantization support in your runner. The selected weights use 1.274 GB; this excludes context cache, runtime buffers and other programs.
These are the first listed tiers that pass our reserve policy, not measured minimum requirements. System RAM and VRAM remain separate. Use the memory tool to change the reserve for your context and runner.