available
Weights
Final and intermediate checkpoints are released.
Inspect this claim’s source ↗Allen Institute for AI (Ai2) / FAMILY RECORD
Olmo 3
A fully open model flow spanning base, mid-trained, long-context, instruction, reasoning, and RL-zero checkpoints.
Checked release: Olmo 3 7B and 32B
Sources checked 2026-09-19
- Olmo 3 model-flow release
- · Primary source
- Olmo 3.1 checkpoint update
- · Primary source
This record begins with the Olmo 3 flow and its cited source also documents the later Olmo 3.1 checkpoint update.
The profile describes this checkpoint, not every release carrying the family name. Developer documentation is not independent reproduction.
THE MODEL / THE TERMS
Know what you are picking up.
- Parameters
- 7B and 32B
- Architecture
- Dense decoder-only transformer with a documented multi-stage training pipeline.
- Modalities
- text
- Context
- Up to 65K tokens in the long-context models
- Weights license
- Apache-2.0 for models and code; data components retain their stated licenses
- Commercial use
- Model and code licenses permit commercial use; downstream users must review dataset licenses.
- Memory
- No universal minimum is asserted; 7B and 32B variants span consumer to multi-GPU deployments.
- Known quantizations
- Community quantizations are available; no single format is treated as canonical here
Specifications are developer-reported. Primary release documentation ↗ · All sources and terms
NINE DIMENSIONS / NO OPENNESS SCORE
What is actually available?
Each component has its own evidence. “Not established” means this source review did not verify availability; it is not a claim that nothing exists.
available
Architecture
Architecture and changes are documented.
Inspect this claim’s source ↗available
Inference code
Model code and deployment instructions are public.
Inspect this claim’s source ↗available
Training code
Pre-training and post-training code are released.
Inspect this claim’s source ↗available
Training recipe
Training stages, configurations, checkpoints, and mixtures are documented.
Inspect this claim’s source ↗available
Data information
Ai2 publishes detailed provenance and composition for Dolma 3 and post-training data.
Inspect this claim’s source ↗available
Training data
Dolma 3 and the released post-training datasets are accessible.
Inspect this claim’s source ↗available
Evaluation
Evaluation suites, data, and scoring standards are public.
Inspect this claim’s source ↗available
Commercial use
Apache-2.0 covers model artifacts; constituent data licenses still apply.
Inspect this claim’s source ↗WHY THIS RELEASE MATTERS
What to explore.
- End-to-end inspectability
- Intermediate checkpoints
- Open data
- Reasoning research
WHAT THE RECORD DOES NOT ESTABLISH
Keep the limits.
- Text-only
- Full reproduction still requires substantial compute
Ai2 documents external research using the open stack, including Goodfire’s post-training behavior tracing; those case studies do not validate every capability claim.
CONNECTED EVIDENCE
A release becomes a starting point.
Thematic or family connections, not proof that an older application uses the selected checkpoint.
FOLLOW THE RECORD
Sources, not superlatives.
- Olmo 3: Charting a path through the model flow ↗ (institution)
- Olmo model flow ↗ (institution)
Source check: 2026-09-19. AI-assisted editorial review, with no named external expert or independent hardware validation claimed. Licenses and model cards can change. Review the linked terms for the specific artifact and intended use.
Understand open source vs open weights ↗