magic incantation to get llama-bench to work with MTP ?

<!-- SC_OFF --><div class="md"><p>currently in order to use MTP one needs to enable it in the starting argument of llama server.</p> <p>--spec-type draft-mtp<br />

<!-- SC_OFF --><div class="md"><p>Anyone got MTP working? It seems like the gguf qants (both unsloth and bartowski) have omitted the mtp tensors</p> </div><!-- SC_ON --> &#32;

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1t3guzw/llamacpp_mtp_support_now_in_beta/"> <img alt="Llama.cpp MTP support now in beta!"

<!-- SC_OFF --><div class="md"><p>So i recently updated LMstudio after a long pause and updated my llama.cpp runtimes too.. i was shocked.. i thought maybe something like

<!-- SC_OFF --><div class="md"><p>this is almost certainly a skill issue, however:</p> <p><code>./llama-bench -hf unsloth/Qwen3.6-27B-GGUF:Q8_0 -sm tensor -ngl 999 -t 1

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1t4uj9h/mtp_on_strix_halo_with_llamacpp_pr_22673/"> <img alt="MTP on strix halo with llama.cpp (PR #22673)"

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1s7z7hj/llamacpp_at_100k_stars/"> <img alt="llama.cpp at 100k stars"

<!-- SC_OFF --><div class="md"><p>I’ve been testing LM Studio on my RTX 5070 Ti (16GB) and Ryzen 9800X3D, running Unsloth Qwen3.5 35B (UD Q4_K_XL).</p> <p>Initially, I thought LM