Recipe
ornith-1-5-35b-a3b-gguf-apex-mtp-i-compact-rtx-3060-12gb-llama-cpp-tp1
ornith-1-5-35b-a3b-gguf-apex-mtp-i-compact-rtx-3060-12gb-llama-cpp-tp1Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.
Record
- Status
- candidate
- Source
- localmaxxing
- Engine
- llama.cpp
- Accelerators
- 1
- Tensor parallel
- 1
- Context tokens
- 65,536
- Max concurrency
- 1
- chat
- unknown
- reasoning
- unknown
- tools
- unknown
- vision
- unknown
Hugging Face model card
Identity
https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF- Repository
- ornith-ai/Ornith-1.5-35B-A3B-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
- Endpoint
http://127.0.0.1:8080- Served model
D:\AI\user\Ornith-1.5-35B-A3B-APEX-MTP-I-Compact.gguf
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status | Sweep |
|---|---|---|---|---|---|---|
| 1 | 65,536 | 1,341.4 | 54.3 | 140.2 | observed | ornith-1-5-35b-a3b-gguf-apex-mtp-i-compact-rtx-3060-12gb-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-3060-12gb
- id
- ornith-1-5-35b-a3b-gguf-apex-mtp-i-compact-rtx-3060-12gb-llama-cpp-tp1
- model instance id
- ornith-ai-ornith-1-5-35b-a3b-gguf--apex-mtp-i-compact
- recipe source
- localmaxxing
- schema version
- local-ai-registry/v1
- speed sweep ids
- ornith-1-5-35b-a3b-gguf-apex-mtp-i-compact-rtx-3060-12gb-llama-cpp-tp1-sweep
- status
- candidate
capabilities
draft launch
- accelerator backend
- nvidia
- arguments
- -hf, ornith-ai/Ornith-1.5-35B-A3B-GGUF, --n-gpu-layers, 999, --host, 0.0.0.0, --port, 8080, -c, 65536
- 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 only | target |
|---|---|
| No | /root/.cache/huggingface |
engine
- name
- llama.cpp
serving
- max concurrency
- 1
- max context tokens
- 65,536
- 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 3060
- localmaxxing · notes
- RTX 3060 12GB, ik_llama.cpp hybrid expert offload: --n-cpu-moe 18 keeps first-18-layer experts in 16GB DDR4, rest full GPU. q8_0 KV, flash-attn, ik build d206417. Sharp chat template + --reasoning-budget 1024. Weights: mudler APEX-MTP I-Compact quant (bundles the native MTP head). Speculative decoding stays OFF here: on A3B-class MoE, draft verification reads the union of per-token expert sets and SLOWS decode; output quality is unchanged either way (lossless verification). Expert cache warmed before measurement.
- localmaxxing · observed command
- # Remote endpoint: http://127.0.0.1:8080 servedModel: D:\AI\user\Ornith-1.5-35B-A3B-APEX-MTP-I-Compact.gguf
- localmaxxing · run id
- cmt9dtiwa00teli011ocu02mp
- localmaxxing · tokenized · endpoint
- 127.0.0.1 ↗
- localmaxxing · tokenized · fidelity
- faithful
- localmaxxing · tokenized · served model name
- D:\AI\user\Ornith-1.5-35B-A3B-APEX-MTP-I-Compact.gguf
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/cmt9dtiwa00teli011ocu02mp ↗ |