AI Benchmark Model

LFM2.5-2.6B

Benchmark scores, pricing, speed, and model comparisons for LFM2.5-2.6B.

Key scores

Intelligence11.0Index score
Coding7.7Index score

LFM2.5-2.6B is a language model developed by Liquid AI and released on August 4, 2026. Within the current evaluation framework, the model holds an intelligence index of 11 and a coding index of 7.7. Performance metrics across specific benchmarks include a score of 0.558 on GPQA, 0.062 on HLE, 0.142 on SciCode, and 0.0533333333333333 on LCR. The model is currently provided at no cost, with input and output pricing listed at $0.00 per million tokens. While specific data regarding output speed and time to first token remains unknown, these benchmark results provide a quantitative overview of the model's capabilities in reasoning, coding, and scientific tasks as of its release date.

Benchmark results

4 tracked benchmark rows for LFM2.5-2.6B.

GPQA Reasoning
55.8%
SciCode Other
14.2%
HLE Other
6.2%
LCR Other
5.3%

Comments (0)

No comments yet

Be the first to share your thoughts!