LFM2.5 350M on Pixel 8 · 8 GB

Experiment Complete Published

llama.cpp b12110 on Pixel 8 · 8 GB, CPU

LFM2.5 350M as Q8_0 (8-bit), archive 2048 input / 128 output: a 2048-token prompt and 128 tokens out. Measured 12 Nov 2026, 11:42 UTC by Glenn Sonna, Xybrid.

Eligible for ranking when the objective's required metrics are measured and this configuration is selected.Approximate comparison: The experiments use different effective KV cache precision: litert-lm default, f16. The experiments here use different 8-bit schemes: litert-int8, Q8_0.

Experiment id s4-pixel8-lfm2.5-350m-8-bit-llama-2048

Summary

Over 6 valid trials of 6 attempted (6 planned), computed from the raw trials below.

MetricMedianRangeMean ± SDQuartiles
Decode speed tok/s, higher is better31.5 30.6–34.5 32.2 ± 1.7 31.2–33.7
Time to first token ms, lower is better10,591 10,155–11,138 10,671 ± 350 10,567–10,893
Peak memory GiB, lower is betterNot measured
Prefill speed tok/s, higher is better193.8 184.3–202.2 192.6 ± 6.4 188.5–194.3

Trials, metric by metric

The line runs from the lowest to the highest valid trial, the dot is the median and each tick is one trial.

Decode speed tok/s, higher is better
Time to first token ms, lower is better
Peak memory GiB, lower is better
Prefill speed tok/s, higher is better

Every trial

Raw token counts and durations, with the speeds computed from them. Invalid trials stay in the record with their reason; they are left out of the summary, never silently dropped.

Every attempted trial
TrialTime to first token msPrefill tok/sDecode tok/sPeak memory GiBThermalNote
111,138184.32,048 tok in 11,112 ms, engine31.3127 tok in 4,063 ms, enginenot recorded37.6 °C, nominal → 37.6 °C, nominal
210,155202.22,048 tok in 10,128 ms, engine34.5127 tok in 3,682 ms, enginenot recorded37.2 °C, nominal → 38.1 °C, nominal
310,587193.92,048 tok in 10,563 ms, engine31.6127 tok in 4,014 ms, enginenot recorded37.6 °C, nominal → 38.3 °C, nominal
410,595193.82,048 tok in 10,567 ms, engine34.4127 tok in 3,695 ms, enginenot recorded37.4 °C, nominal → 38.4 °C, nominal
510,993186.82,048 tok in 10,967 ms, engine30.6127 tok in 4,157 ms, enginenot recorded37.7 °C, nominal → 38.5 °C, nominal
610,560194.42,048 tok in 10,535 ms, engine31.2127 tok in 4,076 ms, enginenot recorded37.4 °C, nominal → 38.8 °C, nominal

Caveats

  • Publisher cards identify the base model but do not always pin the exact export source checkpoint revision.
  • Engines may differ in quantization scheme, effective KV cache precision, kernels, and activations.

Configuration

Device

Device
Google Pixel 8
Cohort
Android phones (mobile)
SoC
Google Tensor G3
CPU
9-core CPU (1× Cortex-X3, 4× Cortex-A715, 4× Cortex-A510)
GPU
Arm Mali-G715
NPU
Not recorded
RAM
8.00 GiB
OS
Android 16 (CP1A.260505.005)
Kernel
6.1.145-android14-11-gfa1d6308d1fe-ab14691759
Power
mains
Low power mode
No
Wakefulness
awake

Engine

Engine
llama.cpp
Maintainer
ggml-org
License
MIT, open source
Version
b12110
Commit
cd959bae9
Binary SHA-256
db82377125f69e87c06b402bba1af072e2a4c4cc9945946e4d6378cc213a84d7
Built from
release
Build flags
None
Backend
CPU
Threads
Not recorded
Interface
http-server
Settings in effect
context_length=4096 parallel=1 speculative=false backend=cpu launch.log_verbosity=4 launch.log_colors=off launch.cache_prompt=false launch.cache_ram_mib=0 launch.cache_reuse=0 launch.continuous_batching=false launch.context_shift=false launch.flash_attn=on launch.cache_type_k=f16 launch.cache_type_v=f16 launch.gpu_layers=0 launch.load_mode=mmap offloaded_layers=0 launch.mmap=true n_batch=2048 n_ubatch=512 n_threads=9 n_threads_batch=9 flash_attn=enabled

Model and weights

Base model
LFM2.5 350M (LFM2.5, 350M)
Upstream
LiquidAI/LFM2.5-350M@9921cd4d82c0e3986887aa4b376f9446729ef8a3
Weights
LiquidAI/LFM2.5-350M-GGUF@657e078c94084481950a2d555a941481f715536b
Format
gguf
File 1
LFM2.5-350M-Q8_0.gguf · 361.7 MiB · sha256 be036a757295e550098b85e13f6af2735d0fa73b41e1156a40c7d8e8e32a5766
File 2
README.md · 3,277 bytes · sha256 4773739eb49540664b16566dc655d03a51df3937639e9dbf28d48a15d01b2436
Quantization
Q8_0 (8-bit)
Weight bits
8
Group size
32
KV cache
f16
Conversion
Published upstream by LiquidAI
Conversion command
Not recorded

Workload

Workload
Archive 2048 input / 128 output
Prompt
2,048 tokens, from prompts/source.txt
Prompt SHA-256
c10d1b7bef62d95322f8f8d229dc66118d9375107bdc49b06f76287d058b2f6c
Chat template
ba551d58630afa3190b1be3602e28301f3d2e9bbac978dfc49d6d825171648b6
Output
128 tokens, ignoring end of sequence
Context length
4,096
Batch size
1
Sampling
Greedy
Timing
One streamed request per trial
Warmup
2 runs, not counted
Trials planned
6
Cooldown
60 s between trials

Conditions

Ambient
Not recorded
Thermal at start
32.3 °C, nominal
Thermal at end
38.8 °C, nominal
Thermal gate
Android thermal status none (0)
Throttling detected
No

Provenance

Measured by
Glenn Sonna, Xybrid
Operator affiliation
Xybrid
Run at
12 Nov 2026, 11:42 UTC
Harness
xybrid-bench 0.4.0 (f109b7085279ab5af52a510133880a070e939eb4)
Methodology
v0.6
Publication
published
Published
16 Nov 2026, 10:23 UTC
Reviewed by
Glenn Sonna

Output check

Result
Passed
Method
sanity-v1-bf16-reference
Reference
checks/sanity-v1.json
Reference SHA-256
a702dee079525ae0ec90465b1c9e26052c99cb3e1c53e83dea05be4c194fc7ca
Required passing checks
7
Check arithmetic
Failed · sha256 4a44dc15364204a80fe80e9039455cc1608281820fe2b24f1e5233ade6af1dd5
Check capital
Passed · sha256 5dd272b4f316b776a7b8e3d0894b37e1e42be3d5d3b204b8a5836cc50597a6b1
Check opposite
Passed · sha256 f151cd9e12c8994edfa316e4624f956bc68460454869177f287146f8cc956a12
Check weekday
Passed · sha256 7d8af1de1262f150187d5938bf649d6c35726970685f916d1ec6392a801b9762
Check color
Passed · sha256 d486dfbd5fb578340bccbdd0a662527eab38550648d5f44517e5ac71b8824703
Check sort
Failed · sha256 edb794d615a63f5eec436d83bed7299d6af36f508c0bf38a3c1cd85d6e4ebd56
Check animal
Passed · sha256 4b227777d4dd1fc61c6f884f48641d02b4d121d3fd328cb08b5531fcacdabf8a
Check word
Failed · sha256 df7e70e5021544f4834bbee64a9e3789febc4be81470df629cad6ddb03320a5c
Check copy
Passed · sha256 d61e6890d0e3571b0c7693612d41f8d059a1fcdef4ba959c0af4144541d41907
Check needle-start
Passed · sha256 8b9af2e5b8b4cfddb87d171997b95c471f8d68a68e30fb5177abeb3332372edd
Check needle-end
Passed · sha256 e6589c736af32f81bf239c5eda6de41959122ccc1d05463b75e0eb0c70fc831c
Check continuation-128
Failed · sha256 d7f917cdf362c04c176faac2a3a39c878a91838b9152de3ff5b965c02ccc8763
Detail
8/8 of the checks the model passes at BF16; needs 7 and the mandatory ones (needle-start, needle-end). Left out, failed by every BF16 configuration: arithmetic, sort, word, continuation-128.

Reproduce it

Commands
pnpm bench run --plan plans/s4-pixel8-lfm2.5-350m-8-bit.json

Three palindromic blocks per quantization class and workload; load, two warmups, one trial, unload per turn.

Raw data