Project AURIONAI Use-Cases Research & Innovation
Allen AI
reasoning

Olmo 3 32B Think

Allen AI's fully open 32B reasoning model. Finetunes 6, quants 33.

Allen AIAllen AIUSAOlmo 3Family
Figure from the Olmo 3 32B Think model card.
Levelmodel
Classreasoning
Parameters32B
Contextnot in the notes
Released15 Dec 2025
Opennessfully open
Runs onsingle datacentre GPU, consumer GPU
Statusactive

From the model's own page; who made each picture and where it comes from are listed on the Image Credits page.

Benchmark Results

vendor-reported
AIME 2025 2025math
vendor-reported
72.5
050100

Score in %. Higher is better.

huggingface.co
IFEvalknowledge and reasoning
vendor-reported
89
050100

Score in %. Higher is better.

huggingface.co
MATHmath
vendor-reported
96.1
050100

Score in %. Higher is better.

huggingface.co
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).

Lineage

finetunesuccessor
Olmo-3-1125-32B to Olmo 3.1 32B Instruct: finetune, high confidenceOlmo-3-1125-32B to Olmo 3 32B Think: finetune, high confidenceOlmo 3 32B Think to Olmo 3.1 32B Think: successor, medium confidenceOlmo-3-1125-32BOlmo-3-1125-32BAllen AIOlmo 3.1 32B InstructOlmo 3.1 32B InstructAllen AI, Dec 2025Olmo 3 32B ThinkOlmo 3 32B ThinkAllen AI, Dec 2025Olmo 3.1 32B ThinkOlmo 3.1 32B ThinkAllen AI
Allen AIOlmo-3-1125-32B is the parent (finetune)high confidencemodel card
Allen AIOlmo 3.1 32B Think is a child (successor)medium confidencenotes
Where this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. The tree shows two generations each way, 4 models. Confidence and evidence are the notes' own.

R&D Topics

Reasoning via large-scale RLproduced by

Identity

Developer
Allen AI
Family
Olmo 3
Released
15 Dec 2025 sourcesnotes
Weights published
–
Record checked
10 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
dense transformer
Total parameters
32B sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
text
Output
text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
fully open
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
single datacentre GPU, consumer GPU
Footprint
Consumer GPU at 4-bit
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Industry Use Cases

No industry use case links this model.

Sources

  1. allenai/Olmo-3-32B-Think , Hugging Face huggingface.co https://huggingface.co/allenai/Olmo-3-32B-Think
  2. allenai (Ai2) huggingface.co https://huggingface.co/allenai/models?search=olmo

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Olmo 3 32B Think", https://aurion.semai.id/models/allen-ai-olmo-3-32b-think/, records of October 2026.