#2531 bench: community results, RTX 5060 Ti 16 GB, 64GB ram- Coder IQ1_M and Swift 1.5 IQ3_XXS

Description

## Hardware and software

- GPU and VRAM: NVIDIA GeForce RTX 5060 Ti, 16 GiB VRAM
- CPU: AMD Ryzen 7 2700X Eight-Core Processor (8 cores, 16 threads); installed RAM: 64.0 GiB; storage: Gen3 SSD
- OS: Windows 10 (10.0.19045); GPU driver: 610.62; CUDA: not read
- Strata: 0.1.44.1 official build


## Results - Qwen3.8-Flash-Next-GSQ-RCO-IQ1_M

| Configuration | Actual prompt tokens | Reused tokens | Generated tokens | Runs | Prompt tok/s median and range | Decode tok/s median and range | TTFT seconds median and range |
| --- | ---: | ---: | ---: | ---: | --- | --- | --- |
| short chat | 46 (43-51) | 0 | 100 (100-100) | 9 | not measured | 44.9 (42.8-53.9) | 0.66 (0.59-0.76) |
| 4K prompt, fresh | 4019 (4019-4019) | 0 (0-0) | 24 cap | 3 | 1235.3 (1228.5-1249.3) | not measured | 3.28 (3.25-3.31) |
| 16K prompt, fresh | 16058 (16058-16058) | 0 (0-0) | 24 cap | 3 | 1666.6 (1663.4-1667.2) | not measured | 9.68 (9.67-9.70) |
| agent session, follow-up turns | 14512 (14510-14547) at the last turn | see .md | up to 64 | 3 | see .md | 57.8 (40.3-70.8) | 3.52 (3.17-3.66) |

## Results - Swift-Qwen3.8-Flash-Next-GSQ-RCO-IQ3_XXS

| Configuration | Actual prompt tokens | Reused tokens | Generated tokens | Runs | Prompt tok/s median and range | Decode tok/s median and range | TTFT seconds median and range |
| --- | ---: | ---: | ---: | ---: | --- | --- | --- |
| short chat | 46 (43-51) | 0 | 100 (100-100) | 9 | not measured | 46.5 (40.5-57.4) | 0.63 (0.57-0.77) |
| 4K prompt, fresh | 4019 (4019-4019) | 0 (0-0) | 24 cap | 3 | 682.1 (678.5-686.3) | not measured | 5.92 (5.91-5.95) |
| 16K prompt, fresh | 16058 (16058-16058) | 0 (0-0) | 24 cap | 3 | 1063.5 (1061.5-1065.1) | not measured | 15.14 (15.12-15.19) |
| agent session, follow-up turns | 14567 (14560-14588) at the last turn | see .md | up to 64 | 3 | see .md | 56.6 (37.7-68.0) | 5.66 (5.28-5.84) |