llamaperf

Step-3.5-Flash

3 reports

Step-3.5-Flash VRAM requirements by size and quant →

How does Step-3.5-Flash run on your hardware?

Pick your GPU or Mac, then read the speeds people reported for Step-3.5-Flash on it, or estimate memory fit and speed.

Free to use. No account needed. Memory estimates and community measurements are labelled separately.

Run Step-3.5-Flash yourself? Add your numbers.

Your coding agent can run a speed test on your machine and send it here with a free account. Or paste a result you've already posted.

Reported performance for Step-3.5-Flash

Filter this model’s reports by setup →

Speeds are as each source reported them. GPU count, offloading and concurrent requests change them, and prompt-processing speed is input throughput, not time to first token. Check the full setup before comparing two reports.

Step-3.5-Flash / b8118 Vulkan / PP512 TG128 at prepared depth8000

AMD Radeon 8060S gfx1151 (Strix Halo; CPU SKU and installed RAM unknown) · llama.cpp

generation:
24.3 tokens/s
prompt processing (prefill):
143.7 tokens/s
quant:
Q3_K_M (GGUF)
kv:
Q8
flash attention:
on
mtp (multi-token prediction):
off

User reports Step-3.5-Flash Q3_K_M at 24.3 t/s generation and 143.7 t/s prompt processing on a single Radeon 8060S gfx1151 device. Setup is llama.cpp build b8118 with Q3_K_M, q8_0 KV cache, -ngl 999, -fa 1 and depth 8000, run on the Vulkan/RADV BLAS arm. The headline pair comes from that same arm, with TG128 24.31 ± 0.81 tok/s and PP512 143.69 ± 1.41 tok/s; the ROCm/HIP arm measured 18.30 ± 1.13 and 158.18 ± 1.89 tok/s.

Oct 13, 2026

Step-3.5-Flash / b8118 Vulkan / PP512 TG128 at prepared depth8000

AMD Radeon 8060S gfx1151 (Strix Halo; CPU SKU and installed RAM unknown) · llama.cpp

generation:
24.6 tokens/s
prompt processing (prefill):
160.5 tokens/s
quant:
IQ3_M (GGUF)
kv:
Q8
flash attention:
on
mtp (multi-token prediction):
off

User reports Step-3.5-Flash IQ3_M at 24.6 t/s generation and 160.5 t/s prompt processing on a single Radeon 8060S gfx1151 device. Setup is llama.cpp build b8118 with IQ3_M, q8_0 KV cache, -ngl 999, -fa 1, -ub 2048 -b 2048 and depth 8000, run on the Vulkan/RADV BLAS arm. The headline pair comes from that arm only, with no cross-pairing; the ROCm/HIP arm measured 16.69 t/s TG128 and 87.66 t/s PP512. PP512 at depth 8000 is not an 8000-token prefill, and TG128 is 128 synthetic timed decode steps.

Oct 13, 2026
Tone: mixedNVIDIA hardware
generation:
17.5 tokens/s
quant:
IQ4_XS (GGUF)
kv:
Q8_0

User reports Step-3.5-Flash at about 17/18 t/s on llama.cpp with an RTX 3090 and two RTX 5060 Ti 16GB cards, 64GB RAM and a Ryzen 9 5950X. Setup is llama.cpp with an IQ4_XS GGUF around 96GB, 16384 context and Q8_0 KV cache, with the model split across GPU and system RAM. The same model on ik_llama.cpp reaches only 10 t/s and is a bit unstable, so the user asks whether the multi-GPU mix is the problem or whether the setup can be refined.

Oct 8, 2026

Get a weekly email of new Step-3.5-Flash reports on any GPU.

Email me new reports

Reported configurations

Individual observations from this page, not expected speeds or a ranking. Select a model to open its report, with the source and offloading conditions. Reported context may be a limit rather than actual input length.

Model / hardwareQuant / engineReported contextGeneration
Step-3.5-Flash 196B (11B active)
3× NVIDIA RTX 3090
IQ4_XS
llama.cpp
16,38417.5 tokens/s