Local AI Registry

Recipe

qwen3-6-27b-q4-k-m-rtx-4090-24gb-llama-cpp-tp1

qwen3-6-27b-q4-k-m-rtx-4090-24gb-llama-cpp-tp1
JSON APIFind compatible recipes

Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.

Record

Status
candidate
Source
localmaxxing
Engine
llama.cpp
Engine version
d1b34251b
Accelerators
1
Tensor parallel
1
Context tokens
76,000
Max concurrency
1
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/unsloth/Qwen3.6-27B-MTP-GGUF
Repository
unsloth/Qwen3.6-27B-MTP-GGUF
Status
known
Link type
Exact Hub repository

Public Hugging Face repository confirmed by the Hub API.

Candidate: useful compatibility or speed evidence. The registry does not offer Run until promotion requirements are met.

Observed configuration

llama.cpp

Evidence only · candidate · reference

Candidate evidence — not a Run contract

Source
https://www.localmaxxing.com/en/runs/cmrdewe850028o401el9ttquj

Observed source tokens

Mechanical split of the source command. Unverified against the engine CLI. The registry does not offer Run for this recipe.

  1. .\llama-server.exe
  2. -m
  3. unsloth\Qwen3.6-27B-MTP-GGUF\Qwen3.6-27B-Q4_K_M.gguf
  4. --spec-type
  5. draft-mtp
  6. --spec-draft-n-max
  7. 3
  8. --threads
  9. 8
  10. --threads-batch
  11. 8
  12. --ctx-size
  13. 76000
  14. --predict
  15. 16384
  16. --ubatch-size
  17. 512
  18. --gpu-layers
  19. all
  20. --flash-attn
  21. on
  22. --no-mmap
  23. --temp
  24. 0.6
  25. --top-k
  26. 20
  27. --min-p
  28. 0.0
  29. --parallel
  30. 1
  31. --metrics
  32. --cache-ram
  33. 16384
  34. --checkpoint-min-step
  35. 512
  36. --no-mmproj
  37. --kv-unified
FlagValue
-munsloth\Qwen3.6-27B-MTP-GGUF\Qwen3.6-27B-Q4_K_M.gguf
--spec-typedraft-mtp
--spec-draft-n-max3
--threads8
--threads-batch8
--ctx-size76000
--predict16384
--ubatch-size512
--gpu-layersall
--flash-attnon
--temp0.6
--top-k20
--min-p0.0
--parallel1
--cache-ram16384
--checkpoint-min-step512

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
176,0002,349101—observedqwen3-6-27b-q4-k-m-rtx-4090-24gb-llama-cpp-tp1-sweep

Remaining fields

Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.

hardware count
1
hardware id
rtx-4090-24gb
id
qwen3-6-27b-q4-k-m-rtx-4090-24gb-llama-cpp-tp1
model instance id
unsloth-qwen3-6-27b-mtp-gguf--q4-k-m
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
qwen3-6-27b-q4-k-m-rtx-4090-24gb-llama-cpp-tp1-sweep
status
candidate

capabilities

draft launch

accelerator backend
nvidia
arguments
-hf, unsloth/Qwen3.6-27B-MTP-GGUF, --n-gpu-layers, 999, --host, 0.0.0.0, --port, 8080, -c, 76000
container port
8,080
environment · LLAMA CACHE
/root/.cache/huggingface
host port
8,080
image
ghcr.io/ggml-org/llama.cpp:server-cuda12-b10481@sha256:b2497f8834f5ecb4e38530f6bf2734b8e0be107f0f0857e259672d1cb85b71c2
ipc
host
kind
docker
shm size
16g
synthesized · generated at
2026-08-31T22:12:17Z
synthesized · image provenance
gemma-4-12b-q4-k-m-rtx-3060-12gb-llama-cpp-tp1
synthesized · template
llama-cpp-server-v1

mounts

read onlytarget
No/root/.cache/huggingface

engine

name
llama.cpp
version
d1b34251b

serving

max concurrency
1
max context tokens
76,000
tensor parallel
1
Provenance & metadata (3)

facts

capabilities.chat · provenance · captured at
2026-08-30T09:10:02Z
capabilities.chat · reason capability-not-verified
capabilities.chat · state unknown
capabilities.reasoning · provenance · captured at
2026-08-30T09:10:02Z
capabilities.reasoning · reason capability-not-verified
capabilities.reasoning · state unknown
capabilities.tools · provenance · captured at
2026-08-30T09:10:02Z
capabilities.tools · reason capability-not-verified
capabilities.tools · state unknown
capabilities.vision · provenance · captured at
2026-08-30T09:10:02Z
capabilities.vision · reason capability-not-verified
capabilities.vision · state unknown
engine.graph mode · provenance · captured at
2026-08-30T09:10:02Z
engine.graph mode · reason runtime-detail-not-published
engine.graph mode · state unknown
serving.kv cache tokens · provenance · captured at
2026-08-30T09:10:02Z
serving.kv cache tokens · reason kv-cache-capacity-not-published
serving.kv cache tokens · state unknown

metadata

localmaxxing · backend
cuda
localmaxxing · hardware label
RTX 4090
localmaxxing · observed command
.\llama-server.exe -m unsloth\Qwen3.6-27B-MTP-GGUF\Qwen3.6-27B-Q4_K_M.gguf --spec-type draft-mtp --spec-draft-n-max 3 --threads 8 --threads-batch 8 --ctx-size 76000 --predict 16384 --ubatch-size 512 --gpu-layers all --flash-attn on --no-mmap --temp 0.6 --top-k 20 --min-p 0.0 --parallel 1 --metrics --cache-ram 16384 --checkpoint-min-step 512 --no-mmproj --kv-unified
localmaxxing · run id
cmrdewe850028o401el9ttquj
localmaxxing · tokenized · arguments
.\llama-server.exe, -m, unsloth\Qwen3.6-27B-MTP-GGUF\Qwen3.6-27B-Q4_K_M.gguf, --spec-type, draft-mtp, --spec-draft-n-max, 3, --threads, 8, --threads-batch, 8, --ctx-size, 76000, --predict, 16384, --ubatch-size, 512, --gpu-layers, all, --flash-attn, on, --no-mmap, --temp, 0.6, --top-k, 20, --min-p, 0.0, --parallel, 1, --metrics, --cache-ram, 16384, --checkpoint-min-step, 512, --no-mmproj, --kv-unified
localmaxxing · tokenized · fidelity
faithful

provenance

captured at
2026-08-30T09:10:02Z

sources

captured atkindurl
2026-08-30T09:10:02Znormalized-recipewww.localmaxxing.com/en/runs/cmrdewe850028o401el9ttquj ↗