AI Tools
OLMo
OLMo : prix, capacités, mises à jour et alternatives, lus sur les pages de l’éditeur.

En anglais
Key facts
- Price
- Not stated on the maker’s pages
- Free plan
- Not stated on the maker’s pages
- Platforms
- Web
- API
- Not stated on the maker’s pages
- Open source
- Yes · Apache-2.0 Read on github.com
- Latest version
- Olmo 3
- Last update
- 7 Oct 2026 Read on github.com
Ai2’s OLMo family includes base, reasoning-focused Think, and chat-oriented Instruct models. The project shares materials from across model development, including training data, code, checkpoints, and evaluation resources. The product page currently presents the Olmo 3 family.
Researchers and developers can inspect, adapt, or run the models for language tasks such as reasoning, coding, and chat. Using them requires suitable software and computing resources; Ai2’s pages do not establish a subscription price or a hosted API price.
- Category
- Developer Tools
- Read from the maker’s pages
- 2026-11-18
Company
Features
- Base, Think, and Instruct variants are listed for the Olmo 3 family.
- Model artifacts include weights, code, and reports.
- Shares pretraining, mid-training, and post-training data.
- Think variants are designed for complex reasoning tasks.
- Instruct variants support chat, tool use, and multi-turn dialogue.
- Provides OlmoCore training tools and open evaluation resources.
- Documentation describes serving models through an OpenAI-compatible endpoint.
Use cases
- Run a language model for chat or question answering.
- Adapt a base model for a specialized research task.
- Study how training data and methods affect model behavior.
- Evaluate model checkpoints with reproducible methods.
- Build and test tool-using language-model applications.
Capabilities
- Self-hosted — “Olmo 3 is a family of compact, dense models at 7 billion and 32 billion parameters that can run on everything from laptops to research clusters.” source
- API — “or use them via API courtesy of our inference partners.” source
- Runs models for you — “or use them via API courtesy of our inference partners.” source
Latest updates
- archive/speed-process-workers
Thread prefetch remains the default for mixed mid-training; process mode is available.
- archive/speed-norms-interval
Gradient-norm bookkeeping now runs once per step.
- archive/speed-no-locks
OLMoOCR pages are rendered in helper processes using thread-local shard handles.
- archive/speed-full-window · OLMo 3.5
The OLMo 3.5 recipes leave prefetch_keep_full off.
- archive/speed-device-once: Stage the pixels to the device through a pinned buffer on a side stream
Image batches are staged through a reusable page-locked buffer and copied on a side stream.
Where it works
Web
Alternatives to OLMo
- Liquid Foundation Models
Liquid Foundation Models are open-weight AI models for text, vision, audio, and decision tasks, usable locally or through Liquid's API.
- Baichuan Open Platform
Baichuan Open Platform provides AI model APIs and tools for building enterprise chat, knowledge-base, and agent applications.
- HHEM (Hughes Hallucination Evaluation Model)
An open model that scores whether generated text is supported by supplied evidence, for checking RAG and AI-agent outputs.
- Baseten
Baseten provides infrastructure to deploy and run AI models, train models, and build agent workflows using managed inference, APIs, training tools, and sandboxes.
- Cerebras Inference
Cerebras Inference serves AI models through a cloud API and web playground, with pay-per-token access for developers and reserved options for enterprises.
- Ollama
Ollama runs open AI models locally or in Ollama’s cloud, with a CLI and API for using them in apps and coding tools.
More from Allen Institute for AI
- Semantic Scholar
Semantic Scholar helps researchers search scientific papers, explore citations, save references, and follow new research.
Questions about OLMo
Which platforms does OLMo run on?
OLMo runs on Web, as its maker’s pages state. Read on the maker’s pages on 18 Nov 2026.
Is OLMo open source?
Yes. The code of OLMo is published at github.com/allenai/Olmo-core under the Apache-2.0 licence. Read on the maker’s pages on 18 Nov 2026.
Is OLMo still available?
Yes, as of 18 Nov 2026: its maker’s pages were read that day and carried no notice that OLMo has ended.
What changed recently in OLMo?
The latest entry in its maker’s changelog: 7 Oct 2026, “archive/speed-process-workers”. Before it: 7 Oct 2026, “archive/speed-norms-interval”. Changelog checked on 18 Nov 2026.
Sources
- https://allenai.org/olmo — Official product page: Olmo 3 family, model variants, artifacts, and model flow.
- https://docs.allenai.org/models/olmo/index.html — Official documentation: model capabilities, licensing, and serving guidance.
- https://allenai.org/blog/olmocore3 — Ai2 announcement dated October 1, 2026, describing Olmo-core 3.
Not confirmed
- Prices
- Search for OLMo pricing did not locate an official pricing page.
- No official pricing page was found to verify tiers or prices.
- Free plan
- The pages describe open model artifacts but do not specify a product pricing tier.
- API
- Documentation describes a locally served compatible endpoint, but does not confirm a hosted OLMo API.