Skip to content
ModelCap

Model decision surface

Compare AI models

Start with Olmo 3 1125 32B and LFM2 8B A1B, or choose any two current ranked language models. Compare capability evidence, price, context, provider availability, and weight access without pretending one field decides every use case.

Current public data

Olmo 3 1125 32B vs LFM2 8B A1B

Live dataset updated 9/12/2026, 1:09:54 PM UTC

Open 1200×630 evidence receipt
Factual comparison of Olmo 3 1125 32B and LFM2 8B A1B
Field
Olmo 3 1125 32B

Allen Institute

LFM2 8B A1B

LiquidAI

ModelCap position#171#170
Index score18.518.6
EvidenceEstimatedpublisher-corpus-prior over 188 held-out anchors (9%); launch card against 3 resolved peers on 6 rows (91%); exceeds every named peer on 1 of 6 rows; no cross-lab optimism probe available; shrunk 0.2 up toward the measured corpusEstimatedglobal-corpus-prior over 188 held-out anchors (19%); launch card against 3 resolved peers on 10 rows (81%); exceeds every named peer on 8 of 10 rows; optimism haircut 9.5 from cross-lab-probe; shrunk 8.1 up toward the measured corpus
Input / 1MUnavailableUnavailable
Output / 1MUnavailableUnavailable
Pricing statusunavailableunavailable
Context——
Providers00
Weight accessOpen weightsRestricted license

Decision facts

  • Olmo 3 1125 32B is #171; LFM2 8B A1B is #170 on the same current language board.
  • Index scores are 18.5 for Olmo 3 1125 32B and 18.6 for LFM2 8B A1B. Their published uncertainty intervals overlap, so the rank difference alone does not establish a reliable capability advantage for your workload.
  • Both positions use Estimated evidence.
  • Weight access differs: Olmo 3 1125 32B is open; LFM2 8B A1B is restricted.

These are separate published fields, not a synthetic winner. ModelCap does not collapse price, access, context, and capability evidence into a hidden recommendation score.

Each comparison page is a permanent, shareable URL with the same live figures as this tool: ModelCap Index position, API pricing, context window, provider count, weight access and every shared benchmark board.