Recipe
minimax-m2-7-ud-q3-k-m-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1
minimax-m2-7-ud-q3-k-m-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.
Record
- Status
- candidate
- Source
- localmaxxing
- Engine
- llama.cpp
- Engine version
- b8870
- Accelerators
- 1
- Tensor parallel
- 1
- Context tokens
- 16,384
- Max concurrency
- 1
- chat
- unknown
- reasoning
- unknown
- tools
- unknown
- vision
- unknown
Hugging Face model card
Identity
https://huggingface.co/unsloth/MiniMax-M2.7-GGUF- Repository
- unsloth/MiniMax-M2.7-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
Observed source tokens
Mechanical split of the source command. Unverified against the engine CLI. The registry does not offer Run for this recipe.
llama-server-mMiniMax-M2.7-UD-Q3_K_M.gguf-ngl999-fa1--no-mmap--mlock-c16384-ctkq8_0-ctvq8_0-b4096-ub4096--parallel2
| Flag | Value |
|---|---|
-m | MiniMax-M2.7-UD-Q3_K_M.gguf |
-ngl | 999 |
-fa | 1 |
-c | 16384 |
-ctk | q8_0 |
-ctv | q8_0 |
-b | 4096 |
-ub | 4096 |
--parallel | 2 |
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status | Sweep |
|---|---|---|---|---|---|---|
| 1 | 16,384 | — | 14 | 4,020.2 | observed | minimax-m2-7-ud-q3-k-m-ryzen-ai-max-plus-395-128gb-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
- ryzen-ai-max-plus-395-128gb
- id
- minimax-m2-7-ud-q3-k-m-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1
- model instance id
- unsloth-minimax-m2-7-gguf--ud-q3-k-m
- recipe source
- localmaxxing
- schema version
- local-ai-registry/v1
- speed sweep ids
- minimax-m2-7-ud-q3-k-m-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1-sweep
- status
- candidate
capabilities
engine
- name
- llama.cpp
- version
- b8870
serving
- max concurrency
- 1
- max context tokens
- 16,384
- 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
- vulkan
- localmaxxing · hardware label
- Ryzen AI Max 395
- localmaxxing · notes
- M5 Strix Halo. MiniMax M2.7 huge MoE at unsloth UD-Q3_K_M (95 GB on disk). 512/512 single-stream, 5 prompts. KV q8_0. ~95 GB consumed of 96 GB GTT cap — need to stop other processes first. mmap on/off both work; mlock-on slightly faster (13.95 vs 12.95 t/s). Active params still substantial despite Q3 quant; throughput limited by raw weight bandwidth.
- localmaxxing · observed command
- llama-server -m MiniMax-M2.7-UD-Q3_K_M.gguf -ngl 999 -fa 1 --no-mmap --mlock -c 16384 -ctk q8_0 -ctv q8_0 -b 4096 -ub 4096 --parallel 2
- localmaxxing · run id
- cmoff9sju000oie041q8cst9q
- localmaxxing · tokenized · arguments
- llama-server, -m, MiniMax-M2.7-UD-Q3_K_M.gguf, -ngl, 999, -fa, 1, --no-mmap, --mlock, -c, 16384, -ctk, q8_0, -ctv, q8_0, -b, 4096, -ub, 4096, --parallel, 2
- localmaxxing · tokenized · fidelity
- faithful
provenance
- captured at
- 2026-08-30T09:10:02Z
sources
| captured at | kind | url |
|---|---|---|
| 2026-08-30T09:10:02Z | normalized-recipe | www.localmaxxing.com/en/runs/cmoff9sju000oie041q8cst9q ↗ |