Model benchmarks

LiquidAI/LFM2.5-350M-GGUF:F16 local LLM performance

As of August 2026, LiquidAI/LFM2.5-350M-GGUF:F16 runs at up to 665.0 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).

llama.cpp
ShareRedditX

Model size

Unknown

Peak speed

665.0 tok/s

Average speed

635.1 tok/s

Min memory

3.4 GB

Max context

65,536 tokens

Avg output / run

3,395 tokens

Avg runtime / run

5s

Avg quality

17.4

Benchmark runs

2

GPUs tested

1

Quality by task

Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 2 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

TaskP5 (low)AvgP95 (high)
Overall15.717.419.0
Agent Workflow17.924.130.3
Code Generation2.63.54.4
Role Play & Narrative24.825.726.6
Research & Analysis13.916.218.4

Performance by hardware and tool

Every hardware/tool/quantization combination LiquidAI/LFM2.5-350M-GGUF:F16 has been benchmarked on, ranked by peak token generation speed. Last updated August 2026.

HardwareToolQuantBest tok/sAvg tok/sMemoryContextQualityRuns
NVIDIA GeForce RTX 3080 Tillama.cpp665.0 tok/s635.1 tok/s3.4 GB65,536 tokens17.42

Frequently asked questions

How fast is LiquidAI/LFM2.5-350M-GGUF:F16 for local inference?
Across 2 community benchmark runs, LiquidAI/LFM2.5-350M-GGUF:F16 reaches up to 665.0 tok/s and averages 635.1 tok/s, with the fastest results on NVIDIA GeForce RTX 3080 Ti.
How much memory does LiquidAI/LFM2.5-350M-GGUF:F16 need?
The leanest observed configuration used about 3.4 GB of memory.
Which tools have been used to run LiquidAI/LFM2.5-350M-GGUF:F16?
Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.