How fast is LFM3-Omni-2B-Q8_0 for local inference?

Across 4 community benchmark runs, LFM3-Omni-2B-Q8_0 reaches up to 140.6 tok/s and averages 136.3 tok/s, with the fastest results on NVIDIA GeForce RTX 3060.

How much memory does LFM3-Omni-2B-Q8_0 need?

The leanest observed configuration used about 2.7 GB of memory (quantizations tested: Q8_0).

Which tools have been used to run LFM3-Omni-2B-Q8_0?

Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.

Model benchmarks

LFM3-Omni-2B-Q8_0 local LLM performance

As of November 2026, LFM3-Omni-2B-Q8_0 runs at up to 140.6 tok/s for local inference (best of 4 community benchmark runs across 1 GPU).

llama.cppQ8_0

ShareReddit X

Model size

Peak speed

140.6 tok/s

Average speed

136.3 tok/s

Avg PP

4540.9 tok/s

Min memory

2.7 GB

Max context

32,768 tokens

Avg output / run

25,114 tokens

Avg runtime / run

3m 14s

Avg quality

37.0

Benchmark runs

GPUs tested

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 4 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

Task	P5 (low)	Avg	P95 (high)
Overall	26.3	37.0	50.6
Agent Workflow	52.3	68.9	86.0
Code Generation	1.8	6.1	13.8
Role Play & Narrative	12.8	32.2	63.4
Research & Analysis	32.4	40.9	49.0

Performance by hardware and tool

Every hardware/tool/quantization combination LFM3-Omni-2B-Q8_0 has been benchmarked on, ranked by peak token generation speed. Last updated November 2026.

Hardware	Tool	Quant	Best tok/s	Avg tok/s	Memory	Context	Quality	Runs
NVIDIA GeForce RTX 3060	llama.cpp	Q8_0	140.6 tok/s	136.3 tok/s	2.7 GB	32,768 tokens	37.0	4

Benchmark runs

All 4 LFM3-Omni-2B-Q8_0 runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Open run page →

Frequently asked questions

How fast is LFM3-Omni-2B-Q8_0 for local inference?: Across 4 community benchmark runs, LFM3-Omni-2B-Q8_0 reaches up to 140.6 tok/s and averages 136.3 tok/s, with the fastest results on NVIDIA GeForce RTX 3060.
How much memory does LFM3-Omni-2B-Q8_0 need?: The leanest observed configuration used about 2.7 GB of memory (quantizations tested: Q8_0).
Which tools have been used to run LFM3-Omni-2B-Q8_0?: Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.