Model benchmarks

liquidai/lfm2.5-2.6b local LLM performance

As of August 2026, liquidai/lfm2.5-2.6b runs at up to 165.9 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).

LM Studio
ShareRedditX

Model size

2.6B

Peak speed

165.9 tok/s

Average speed

132.7 tok/s

Min memory

n/a

Max context

32,768 tokens

Avg runtime / run

3m 31s

Avg quality

56.8

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)
Overall53.956.859.8
Agent Workflow81.381.681.9
Code Generation16.717.317.9
Role Play & Narrative63.064.766.4
Research & Analysis52.763.774.8

Performance by hardware and tool

Every hardware/tool/quantization combination liquidai/lfm2.5-2.6b has been benchmarked on, ranked by peak token generation speed. Last updated August 2026.

HardwareToolQuantBest tok/sAvg tok/sMemoryContextQualityRuns
NVIDIA GeForce RTX 4070 Ti SUPERLM Studio—165.9 tok/s132.7 tok/sn/a32,768 tokens56.82

Benchmark runs

All 2 liquidai/lfm2.5-2.6b runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Frequently asked questions

Is lmstudio-community/lfm2.5-2.6b good for coding?
In our benchmarks, lmstudio-community/lfm2.5-2.6b scores 17.3/100 for coding. It runs at about 132.7 tok/s, so if you want more speed, Qwen3.6-35B-A3B-Q4_K_XL is faster (~156.9 tok/s) and still scores well for coding (82.2/100).
Is lmstudio-community/lfm2.5-2.6b good for agentic (tool-using) tasks?
In our benchmarks, lmstudio-community/lfm2.5-2.6b scores 81.6/100 for agentic workflows. It runs at about 132.7 tok/s, so if you want more speed, Tiel-Coder-35B-A3B-Q4_K_S is faster (~162.4 tok/s) and still scores well for agentic workflows (87.8/100).
How fast is liquidai/lfm2.5-2.6b for local inference?
Across 2 community benchmark runs, liquidai/lfm2.5-2.6b reaches up to 165.9 tok/s and averages 132.7 tok/s, with the fastest results on NVIDIA GeForce RTX 4070 Ti SUPER.
Which tools have been used to run liquidai/lfm2.5-2.6b?
Benchmarks were submitted using LM Studio. Results are community-contributed and updated as new runs arrive.