# Zen 6 throughput benchmark · Zen 6 — Hanzo AI > Zen 6, the latest open-weight generation, on three machines Hanzo owns: prompt prefill and single-request decode, in tokens per second, one machine each. [Benchmarks](https://hanzo.ai/benchmarks) / [Zen 6](https://hanzo.ai/benchmarks?product=Zen%206#all) / Zen 6 throughput Zen 6 · Throughput on Hanzo hardware # Zen 6 throughput Zen 6, the latest open-weight generation, on three machines Hanzo owns: prompt prefill and single-request decode, in tokens per second, one machine each. ## Latest [50.6 tok/s decoding code · 112 tok/s across 8 requestsAlso measured · Zen 6 · NVIDIA Grace-Blackwellno baseline · Unscored · Sep 23, 2026 · Hanzo-measured](#history) [262K context, in two slotsAlso measured · Zen 6 · AMD Strix Halono baseline · Unscored · Sep 23, 2026 · Hanzo-measured](#history) [~43 tok/sDecode throughput · Zen 6 · NVIDIA Grace-Blackwellno baseline · Unscored · Sep 23, 2026 · Hanzo-measured](#history) [~2,400 tok/sPrefill throughput · Zen 6 · NVIDIA Grace-Blackwellno baseline · Unscored · Sep 23, 2026 · Hanzo-measured](#history) ## History 8 results across 3 metrics, newest first. Superseded runs stay, so a revision's progress can be read. Measured Subject Metric Value Baseline Result Status Sep 23, 2026 Zen 6 · NVIDIA Grace-Blackwell Also measured 50.6 tok/s decoding code · 112 tok/s across 8 requests – Unscored Frozen · Hanzo-measured Sep 23, 2026 Zen 6 · AMD Strix Halo Also measured 262K context, in two slots – Unscored Frozen · Hanzo-measured Sep 23, 2026 Zen 6 · NVIDIA Grace-Blackwell Decode throughput ~43 tok/s – Unscored Frozen · Hanzo-measured Sep 23, 2026 Zen 6 · AMD Strix Halo Decode throughput ~34 tok/s – Unscored Frozen · Hanzo-measured Sep 23, 2026 Zen 6 · Apple silicon Decode throughput 64–81 tok/s – Unscored Frozen · Hanzo-measured Sep 23, 2026 Zen 6 · NVIDIA Grace-Blackwell Prefill throughput ~2,400 tok/s – Unscored Frozen · Hanzo-measured Sep 23, 2026 Zen 6 · AMD Strix Halo Prefill throughput 1,150–1,400 tok/s – Unscored Frozen · Hanzo-measured Sep 23, 2026 Zen 6 · Apple silicon Prefill throughput 500–690 tok/s – Unscored Frozen · Hanzo-measured ## Conditions NVIDIA Grace-Blackwell: 121 GB unified, NVFP4 weightsAMD Strix Halo: 128 GB unified, 4-bit buildApple silicon: unified memory, MLX builddecode is one request unless a row says otherwisea range is what the runs spread over; the low end is the figure ordered and drawn ## Hardware NVIDIA Grace-Blackwell · AMD Strix Halo · Apple silicon ## Source [Zen 6, measured by Hanzo on Sep 23, 2026](https://hanzo.ai/zen) ## Reproduce No public command regenerates this result. The source above is the record it is read from. ## More Zen 6 benchmarks [Inference vs llama.cpp](https://hanzo.ai/benchmarks/inference) [Every result for Zen 6 throughput on the shelf](https://hanzo.ai/benchmarks?q=Zen%206%20throughput#all) · [How these are measured](https://hanzo.ai/benchmarks#methods)