MMODELYST
Models/LFM2.5-8B-A1B
MDL

LFM2.5-8B-A1B

Liquid AI

Verified against Artificial Analysis · Jun 15, 2026
Capability
40
percentile index
Price /M
$0 / $0
in / out
Context
Avg score
20.3
7 benchmarks
AA Intelligence
14.2
index
Speed
227 tok/s
TTFT 1.8s
Released
May 28, 2026
About

LFM2.5-8B-A1B is a proprietary model in the LFM2.5 family from Liquid AI. Benchmarked on 7 evals, averaging 20.3.

Benchmark scores · 7
avg 20.3 — higher is better
IFBench
55.6
GPQA Diamond
51.3
τ²-bench
16.1
SciCode
7.8
Humanity's Last Exam
6.9
Terminal-Bench Hard
4.5
AA Long-Context Reasoning
0
Details
PublisherLiquid AI
FamilyLFM2.5
Parameters
License
WeightsClosed
ReleasedMay 28, 2026
Performance & price
Output speed227 tok/s
Latency (TTFT)1.8s
First answer11s
Price in / out$0 / $0 per 1M
Coding index5.6

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

Capability over time
τ²-benchfirst seen at 16.1
yesterday
Humanity's Last Examfirst seen at 6.9
6d ago
SciCodefirst seen at 7.8
6d ago
GPQA Diamondfirst seen at 51.3
6d ago
Terminal-Bench Hardfirst seen at 4.5
6d ago
6d ago

Append-only ledger — every observed change to this model's numbers.

API providers
Artificial Analysis$0/M