Skip to main content
Back to Explore

LFM2.5-8B-A1B

Liquid AI·Released 2026-05-28

Performance

219

tokens/sec

Faster than 91% of models

2.16

seconds

Faster than 24% of models

11.30

seconds

Faster than 37% of models

Market Median

92 tok/s

138% faster

Median TTFT

1.13s

92% slower

Speed Comparison

Nova 2.0 Omni (Non-reasoning)
219 tok/s+0%
Google: Gemini 3.5 Flash
220 tok/s+0%
Qwen3 0.6B (Non-reasoning)
222 tok/s+1%

Benchmarks

MMLU-ProNot evaluated
GPQA Diamond
51.3%
HLE
6.9%
LiveCodeBenchNot evaluated
SciCode
7.8%
TerminalBench Hard
4.5%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025Not evaluated
IFBench
55.6%
Long Context Recall
0.0%
Tau2
16.1%
Market AverageTop Score

Quick Compare

Similar Models

Compare all 7 models