Benchmark Almanac

LFM2.5-2.6B

LFM2.5-2.6B places in the top three on ToolSandbox.

Liquid AI Released August 4th, 2026 Family LFM2.5 The lab's model card Compare with others
12
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
Liquid AI Aug 2026

LFM2.5-2.6B

Open weights Small

128.9 #284 of 404
Self-Reported Index, from the lab's own numbers. No Epoch Capabilities Index score yet.
  • IFBench 59.2% #57 /80
  • BFCL V4 56.9% #23 /30
  • ToolSandbox 77.8% #3 /5

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
Multi-IF 80.1% #5 of 28 Lab model card
IFBench 59.2% #57 of 80 Lab model card
AA-Omniscience -29.5 #10 of 10 Lab model card
IFStruct 85.5% #2 of 2 Lab model card

Math and science

BenchmarkScoreRankSource
AIME 2025 51.9% #130 of 143 Lab model card

Coding

BenchmarkScoreRankSource
LiveCodeBench v6 59.4% #75 of 90 Lab model card

Agents and tools

BenchmarkScoreRankSource
ToolSandbox 77.8% #3 of 5 Lab model card
BFCL V4 56.9% #23 of 30 Lab model card
Claw-Eval 62.9% #11 of 13 Lab model card
τ³-bench Banking 5.67% #22 of 22 Lab model card
PinchBench 68.2% #14 of 14 Lab model card
BrowseComp-Plus 26.9% #4 of 4 Lab model card

Its lab's numbers put it at 128.9, number 284 of 404

The Self-Reported Index fits one score to the 8 capability results Liquid AI reported for this model, on the scale of the Epoch Capabilities Index. Labs choose which benchmarks to show and how to run them, so read it as the lab's claim, not an independent measurement. The 90% interval runs from 126.0 to 131.9. How we compute it.

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.