LFM2.5-8B-A1B benchmarks, price, and specs | Tokens or Towers

AA Intelligence: 7.2#130 of 151
Price: Self-host
Output speed: —
Context window: 33K tokens

Composite indexes

Artificial Analysis Intelligence Index#130 of 151
7independentIntelligence Index v4.3; reasoning; estimatedArtificial Analysis: LFM2.5-8B-A1B ↗
Leader: Claude Opus 5.5 at 58

Agents and tool use

τ²-Bench#72 of 89
16.1%independenttelecom; reasoningArtificial Analysis: LFM2.5-8B-A1B ↗
Leader: GLM-5.2 at 99.1%

Reasoning and knowledge

GPQA Diamond#127 of 144
51.3%independentreasoningArtificial Analysis: LFM2.5-8B-A1B ↗
Leader: GPT-6 Astra at 96.1%
Humanity’s Last Exam#117 of 153
6.9%independentno tools; reasoningArtificial Analysis: LFM2.5-8B-A1B ↗
Leader: Claude Sonnet 5.5 at 64.5%
CritPt#134 of 144
0%independentreasoningArtificial Analysis: LFM2.5-8B-A1B ↗
Leader: Muse Wonder Full at 33%

Vision and long context

AA Long Context Reasoning#140 of 144
0%independentreasoningArtificial Analysis: LFM2.5-8B-A1B ↗
Leader: Kimi K3 at 88.7%

Other published results

Result	Score	Reported by	Source
Terminal-Bench Hard (AA) · reasoning	4.5%	independent	Artificial Analysis: LFM2.5-8B-A1B ↗
IFBench · reasoning	55.6%	independent	Artificial Analysis: LFM2.5-8B-A1B ↗
AA-Omniscience Index · reasoning	-33.2 index	independent	Artificial Analysis: LFM2.5-8B-A1B ↗