Back to Explore
LFM2.5-8B-A1B
Liquid AI·Released 2026-05-28
Comparison data ready48% coverage9/9 fields directly observedUpdated Jul 25, 2026, 6:01 AM
Related Models
Performance
332
tokens/sec
Faster than 97% of models
1.49
seconds
Faster than 40% of models
7.52
seconds
Faster than 45% of models
Market Median
95 tok/s
248% faster
Median TTFT
1.19s
25% slower
Speed Comparison
gpt-oss-120b (low)
324 tok/s-2%
Nemotron 3 Nano Omni 30B A3B Reasoning
323 tok/s-3%
Step 3.5 Flash
322 tok/s-3%
Benchmarks
MMLU-ProNot evaluated
GPQA Diamond
46.6%
HLE
6.9%
LiveCodeBenchNot evaluated
SciCode
7.8%
TerminalBench Hard
4.5%
MATH-500
90.2%
AIMENot evaluated
AIME 2025Not evaluated
IFBench
53.3%
Long Context Recall
0.0%
Tau2
16.1%
Market AverageTop Score