Recipe

gpt-oss-120b-mxfp4-intel-arc-pro-b70-32gb-llama-cpp-tp2

gpt-oss-120b-mxfp4-intel-arc-pro-b70-32gb-llama-cpp-tp2

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

Record

Status
candidate
Source
localmaxxing
Engine
llama.cpp
Engine version
mainline 788e07d (2026-07-17)
Accelerators
2
Tensor parallel
2
Context tokens
2,048
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/openai/gpt-oss-120b
Repository
openai/gpt-oss-120b
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/cmrws0j0500i4nz01jduaji7l

No tokenized launch fields. This is measured or documented compatibility, not a Docker launch.

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
12,0481,124.759.4observedgpt-oss-120b-mxfp4-intel-arc-pro-b70-32gb-llama-cpp-tp2-sweep

Remaining fields

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

hardware count
2
hardware id
intel-arc-pro-b70-32gb
id
gpt-oss-120b-mxfp4-intel-arc-pro-b70-32gb-llama-cpp-tp2
model instance id
openai-gpt-oss-120b--mxfp4
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
gpt-oss-120b-mxfp4-intel-arc-pro-b70-32gb-llama-cpp-tp2-sweep
status
candidate

capabilities

engine

name
llama.cpp
version
mainline 788e07d (2026-07-17)

serving

max context tokens
2,048
tensor parallel
2
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
serving.max concurrency · provenance · captured at
2026-08-31T23:03:15Z
serving.max concurrency · reason server-capacity-not-evidenced
serving.max concurrency · state unknown

metadata

localmaxxing · backend
vulkan
localmaxxing · batch size
1
localmaxxing · hardware label
Intel Arc Pro B70
localmaxxing · notes
117B MoE (gpt-oss-120b) MXFP4 59.02 GiB across dual B70, -sm layer -fa 1 -ngl 99. llama-bench r=5: tg128 59.44 +/- 1.15, pp2048 1124.67 +/- 98.37 (pp variance from cold-load effects across reps; treat as ~1100). Notably faster than published dual-B70 120B-MoE dataset figures (~43 t/s). Mainline llama.cpp - no patches needed for this model. Same rig/series as our Nemotron-75B/Qwen3.6-35B/27B/Coder-30B entries; SR-IOV VFs, no power telemetry. peakVramGb approximate (model 59GiB + buffers, near the 63.6GB combined ceiling).
localmaxxing · run id
cmrws0j0500i4nz01jduaji7l

provenance

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

sources

captured atkindurl
2026-08-30T09:10:02Znormalized-recipewww.localmaxxing.com/en/runs/cmrws0j0500i4nz01jduaji7l