Builds / RTX 5060 Ti 16GB
RTX 5060 Ti 16 GB: KRAFTON Raon-1 mini preview IQ4_XS with DFlash2 drafter at 42.1 tok/s, 256K window, 9 layers of experts in RAM
Last updated:
- GPU(s)
- RTX 5060 Ti 16GB
Amazon · eBay (used) · Newegg - VRAM / unified memory (GB)
- 16
- CPU
- not stated
- Motherboard
- not stated
- RAM
- not stated
- PSU
- not stated
- Other
- KRAFTON llama.cpp fork (branch raon-1-mini-preview, fb895662) built from source with CUDA 13.3 (sm_120a); -ngl 999 --n-cpu-moe 9; 262,144-token window, 1 slot; second GPU in the box not used
- Reported total cost
- not stated
- Power
- not stated
Measured speed (generation tokens/sec)
| Model | Quant | Engine | tok/s | Prompt processing (tok/s) | Details |
|---|---|---|---|---|---|
| Raon-1 mini Preview | IQ4_XS GGUF (KRAFTON, 14.97 GiB) + DFlash2 Q8_0 drafter (699 MiB) | llama.cpp KRAFTON fork fb895662 (CUDA 13.3) | 42.1 | 586 @ 8K-token prompt est. TTFT @8k ≈ 14.0 s (calculated from measured prefill) | ⚡ spec-decode (DFlash2)context: short code + prose prompts, 512 generated, 256K window📏 Measured 22 Oct 2026: mean of code 39.9 and prose 44.3 (median of 2 runs each), card's sampling settings, thinking off; 17.7% draft acceptance, mean accepted length 2.2. Long prompts (11 Nov): code/prose decode 39.3/46.4 @16K, 37.8/48.3 @32K, 35.3/45.1 @64K, 25.9/31.3 @128K |
Prompt processing = prefill speed (how fast the prompt is read) as reported by the source, with the prompt length it was measured at; blank = the source does not report it. est. TTFT @8k is calculated from the measured prefill speed (8,192 ÷ prefill tok/s), shown only when prefill was measured on a prompt of 512+ tokens. It is not a measurement and ignores load time and caching. measured TTFT is the source's own time-to-first-token. Why prefill matters for coding assistants.
Independent run using the maker's server command (niklaslenz.com results v2, measured 22 Oct, long prompts 11 Nov 2026). The maker's example uses Q4_K_M (18.33 GiB), which does not fit 16 GB, so the smallest IQ4_XS file was used and the experts of 9 layers sit in system RAM. KRAFTON's own DFlash2 speed-ups (up to 2.18x) are for a B300 with vLLM and are not comparable. No no-drafter baseline was published for this card.
- 1× RTX 5060 Ti 16GB: ≈€739 each (≈$828), new, approx. price as of 2026-10-16, lowest listing on billiger.de (Asus Dual GeForce RTX 5060 Ti OC 16 GB GDDR7)
New retail prices in Germany incl. VAT; used cards are often much cheaper. Prices change daily.
Blank fields = not stated in the source. We never estimate build data; GPU street prices above are separately sourced and dated.
Quick answers: Can RTX 5060 Ti 16GB run Raon-1 mini Preview?
Speculative-decoding rows on this page are not like-for-like with plain decode. See how much MTP and DFlash really speed things up (measured pairs).