# Ai2 — Introducing Olmo-core 3: Open, scalable training infrastructure for large MoEs

- Company: Ai2 (allenai.org)
- Announced: 2026-10-01
- Category: developer-tool-release
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://allenai.org/blog/olmocore3
- Record: https://forck.live/items/18530-introducing-olmo-core-3-open-scalable-training-infrastructure-for-large-moes
- Subject: Olmo / Molmo / Asta

Ai2 released Olmo-core 3, an upgrade to its open-source framework for training large language models, featuring a redesigned mixture-of-experts (MoE) training system. The framework is designed to scale MoE training into the trillion-parameter range and includes techniques such as expert parallelism, pipeline parallelism, and a distributed optimizer. In benchmarks, a 47-billion-parameter MoE achieved 2.7× the throughput of the previous implementation, and MXFP8 support improved training throughput by about 21% over BF16.

## Evidence

Verbatim from https://allenai.org/blog/olmocore3:

> Olmo-core 3 is designed to scale MoE training into the trillion-parameter range while preserving computational efficiency.

---

Record: https://forck.live/items/18530-introducing-olmo-core-3-open-scalable-training-infrastructure-for-large-moes
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md