Qwen3.5-4B has 4.206 billion total parameters. The selected Q4_K_M GGUF file is 2.741 GB.
Data snapshot: . 6 published tests in this catalog.
Model file and details
Provider
Qwen
Total parameters
4,205,751,296
Exact file size
2,740,937,888 bytes
Quantization / precision
Q4_K_M
Weight repository
unsloth/Qwen3.5-4B-GGUF
Original model repository
Qwen/Qwen3.5-4B
Weight revision
e87f176479d0855a907a41277aca2f8ee7a09523
Model revision
851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a
Runtime notes
Confirm model architecture and quantization support in your runner.
Recorded release date
2026-02-27
Date evidence
The official Apache-2.0 source checkpoint was created on 27 February 2026; Unsloth published the complete GGUF repository and selected Q4_K_M artifact on 2 March 2026.
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-Pro
79.1 / 100
Knowledge and reasoning across many subjects. Published accuracy; different evaluation settings can change the result.
Benchmark efficiency: 28.859 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: Qwen publisher evaluation of the official source checkpoint; model-card table row MMLU-Pro. The card does not disclose a pinned evaluator revision, hardware, precision or per-row prompt settings.
Publisher source-checkpoint point estimate, not an independent result and not a benchmark of the selected Q4_K_M file. Quantization, thinking mode, prompt and inference settings can change results.
Hard science questions. This is a science reasoning test, not an overall intelligence score.
Benchmark efficiency: 27.801 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: Qwen publisher evaluation of the official source checkpoint; model-card table row GPQA Diamond. The card does not disclose a pinned evaluator revision, hardware, precision or per-row prompt settings.
Publisher source-checkpoint point estimate, not an independent result and not a benchmark of the selected Q4_K_M file. Quantization, thinking mode, prompt and inference settings can change results.
Solving coding problems. Version 6 only; published pass rates can vary with generation settings.
Benchmark efficiency: 20.358 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: Qwen publisher evaluation of the official source checkpoint; model-card table row LiveCodeBench v6. The card does not disclose a pinned evaluator revision, hardware, precision or per-row generation settings.
Publisher source-checkpoint point estimate, not an independent result and not a benchmark of the selected Q4_K_M file. Quantization, thinking-token budget, prompt and inference settings can materially change coding results.
Following precise instructions. These cards report IFEval without consistently specifying prompt-level versus instruction-level scoring. Treat this view as approximate.
Benchmark efficiency: 32.763 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: Qwen publisher evaluation of the official source checkpoint; model-card table row IFEval. The card does not disclose a pinned evaluator revision, hardware, precision or per-row prompt settings.
Publisher source-checkpoint point estimate, not an independent result and not a benchmark of the selected Q4_K_M file. Quantization, thinking mode, prompt and inference settings can change results.
Choosing and calling functions. Published BFCL v4 scores; tool environments and settings differ.
Benchmark efficiency: 18.351 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: Qwen publisher evaluation of the official source checkpoint; model-card table row BFCL-V4. The card does not disclose a pinned evaluator revision, hardware, precision, tool parser or per-row generation settings.
Publisher source-checkpoint point estimate, not an independent result and not a benchmark of the selected Q4_K_M file. Tool schemas, parser behavior, quantization and runtime configuration can change function-calling results. The 2.741 GB language-weight file also excludes the vision projector, runtime, KV cache and context memory and does not establish device speed or compatibility.
Revised MMLU knowledge test. Kept separate from original MMLU and MMLU-Pro.
Benchmark efficiency: 32.398 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: Qwen publisher evaluation of the official source checkpoint; model-card table row MMLU-Redux. The card does not disclose a pinned evaluator revision, hardware, precision or per-row prompt settings.
Publisher source-checkpoint point estimate, not an independent result and not a benchmark of the selected Q4_K_M file. Quantization, thinking mode, prompt and inference settings can change results.
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 2.741 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 2.74 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 2.741 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.