Models

Loading community benchmarks…

← All models

Community benchmarks

Qwen3 8B

Submitted benchmarks for Qwen3 8B. Browse individual runs and their measurement settings, or open the report JSON for full benchmark details.

Compare chips Compare runtimes Add your benchmark

Submitted runs: 65
Chips tested: 9
Contributors: 3

Model identity

Grouped as Qwen/Qwen3-8B. The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution. · This report predates artifact identity, so ComputeArena mapped its model name to a model family by hand. The exact model bytes were not verified.. Artifact verification confirms published model bytes, not that a community-submitted benchmark was honestly executed. Hover or tap a label for what it means.

This model group collects runs reported as basecompute/Qwen3-8B, Qwen3-8B, Qwen3-8B-base-q2, Qwen3-8B-base-q3, Qwen3-8B-base-q5, Qwen3-8B-base-q6, Qwen3-8B-Q4, Qwen3-8B-Q8. Individual results keep the package name and format they were submitted with.

View on Hugging Face ↗

Published files benchmarked

basecompute/Qwen3-8B/Qwen3-8B-Q4.base (opens in a new tab)BaseRTQ4· 12 runs
basecompute/Qwen3-8B/Qwen3-8B-Q8.base (opens in a new tab)BaseRTQ8· 6 runs
unsloth/Qwen3-8B-GGUF/Qwen3-8B-Q4_K_M.gguf (opens in a new tab)llama.cppQ4_K_M· 4 runs
unsloth/Qwen3-8B-GGUF/Qwen3-8B-BF16.gguf

Benchmark results

Peak throughput is the highest reported value and may come from different runs. Compare model, quantisation, backend, and token counts before drawing conclusions. Model verification identifies published artifact bytes; community benchmark execution is not remotely attested.

Individual runs for Qwen3 8B. Throughput is in tokens per second; dates use your browser timezone. Hover a date for the exact local time.
Model / runtime / format	Chip / backend	Decode	Prefill	Peak process memory	Contributor

Qwen3-8B llama.cppBF16 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	43.5 tok/s TG128	3,133 tok/s PP512	17,455.2 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ8_0 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	64.0 tok/s TG128	3,207 tok/s PP512	12,148.5 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ8_0 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	74.2 tok/s TG128	3,241 tok/s PP512	10,131.9 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ6_K The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	80.1 tok/s TG128	2,903 tok/s PP512	8,973.3 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ6_K The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	84.0 tok/s TG128	2,823 tok/s PP512	8,240.2 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ5_K_M The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	86.8 tok/s TG128	2,869 tok/s PP512	7,431.7 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ5_K_M The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	87.5 tok/s TG128	2,837 tok/s PP512	7,405.6 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ4_1 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	107.3 tok/s TG128	2,894 tok/s PP512	6,830.6 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ4_K_M The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	93.2 tok/s TG128	3,076 tok/s PP512	6,723.7 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ4_K_M The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	93.7 tok/s TG128	3,110 tok/s PP512	6,620.7 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ3_K_M The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	91.6 tok/s TG128	2,911 tok/s PP512	5,937.2 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ3_K_M The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	91.6 tok/s TG128	2,864 tok/s PP512	5,762.8 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ2_K The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	105.5 tok/s TG128	2,834 tok/s PP512	5,169.5 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ2_K The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	AMD Radeon RX 7900 XT ROCm	112.3 tok/s TG128	2,798 tok/s PP512	4,959.3 MiB	arki05	11 Sep 2026	View Benchmark
basecompute/Qwen3-8B BaseRTQ8· default-q8 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	Apple M5 Pro Metal	34.5 tok/s TG128	1,744 tok/s PP512	10,376.0 MiB	arki05	11 Sep 2026	View Benchmark
basecompute/Qwen3-8B BaseRTQ4· default-q4 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	Apple M5 Pro Metal	60.7 tok/s TG128	1,755 tok/s PP512	6,269.8 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B-base-q6 BaseRTQ6 This report predates artifact identity, so ComputeArena mapped its model name to a model family by hand. The exact model bytes were not verified.	Apple M5 Pro Metal	42.1 tok/s TG128	1,723 tok/s PP512	8,448.1 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B-base-q5 BaseRTQ5 This report predates artifact identity, so ComputeArena mapped its model name to a model family by hand. The exact model bytes were not verified.	Apple M5 Pro Metal	48.4 tok/s TG128	1,730 tok/s PP512	7,353.1 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B-base-q3 BaseRTQ3 This report predates artifact identity, so ComputeArena mapped its model name to a model family by hand. The exact model bytes were not verified.	Apple M5 Pro Metal	62.0 tok/s TG128	1,748 tok/s PP512	5,710.4 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B-base-q2 BaseRTQ2 This report predates artifact identity, so ComputeArena mapped its model name to a model family by hand. The exact model bytes were not verified.	Apple M5 Pro Metal	81.8 tok/s TG128	1,786 tok/s PP512	4,615.7 MiB	arki05	11 Sep 2026	View Benchmark
basecompute/Qwen3-8B BaseRTQ4· default-q4 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	Apple M5 Pro Metal	59.7 tok/s TG128	1,761 tok/s PP512	6,206.0 MiB	isu	11 Sep 2026	View Benchmark
basecompute/Qwen3-8B BaseRTQ4· default-q4 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	Apple M5 Pro Metal	58.2 tok/s TG128	1,758 tok/s PP512	6,266.1 MiB	isu	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppBF16 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	Apple M5 Pro BLAS + Metal	18.9 tok/s TG128	1,376 tok/s PP512	18,109.9 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ8_0 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	Apple M5 Pro BLAS + Metal	29.2 tok/s TG128	1,448 tok/s PP512	12,808.3 MiB	arki05	11 Sep 2026	View Benchmark
Qwen3-8B llama.cppQ8_0 The report's model SHA-256 matches a file published on Hugging Face, so the exact model bytes are known. This does not attest benchmark execution.	Apple M5 Pro BLAS + Metal	34.4 tok/s TG128	1,473 tok/s PP512	10,786.9 MiB	arki05	11 Sep 2026	View Benchmark