Long Context & Efficient Sequences · 2025
MiniMax-01: Scaling Foundation Models with Lightning Attention
First large-scale (456B-parameter MoE) production model built on linear (lightning) attention, demonstrating million-token training context and 4M-token inference at frontier quality.
Editorial record
Plain-language summary
First large-scale (456B-parameter MoE) production model built on linear (lightning) attention, demonstrating million-token training context and 4M-token inference at frontier quality.
Knowledge graph
Relationships
Descendants
Depends onEvidence: Strongly supported
Attention Is All You Need
Builds on the long context lineage in the archive
freshness sweep 2026
Source record
Provenance
- Record ID
- P-609
- Record created
- 2026-07-13
- Last reviewed
- 2026-07-14
- Record version
- 2
- https://arxiv.org/abs/2501.08313
- arXiv:2501.08313
Citation caveat: Citation metadata is approximate and marked unverified in the source dataset.