LINEAiGE
AI HISTORY

Aya Vision Public Launch

Cohere Labs announced Aya Vision on March 4, 2025 as a multilingual multimodal model supporting text-and-image tasks across 23 languages.

4 March 2025AI HISTORYVERIFIED

Historical classification

This verified record belongs to LINEAiGE's artificial-intelligence history layer beginning in 1956.

Why it matters

A dated public-launch milestone in the Aya research program, extending the program into multilingual multimodal text-and-image capabilities.

People, organizations & technologies

ORGANIZATIONS
Cohere Labs

TECHNOLOGIES
Aya Vision, Aya Vision 32B

Evidence & verification

STATE
verified

CONFIDENCE
high

LAST VERIFIED
2026-09-24T17:06:18+08:00

Cohere Labs' first-party launch post is dated March 4, 2025 and announces Aya Vision as an open-weights multilingual multimodal model. Cohere's first-party model documentation independently corroborates the Aya Vision 32B model, text-and-image modality, 23-language scope, and open-weight availability. The later primary technical report, submitted May 13, 2025, documents Aya Vision 8B and 32B and is retained as chronology context rather than evidence for the March launch date. This candidate records the dated public-launch milestone only and does not infer development geography, ancestry, downstream influence, causality, or relationship edges.

Sources

SOURCE 1 · PRIMARY · CLAIM-EVIDENCE

Aya Vision: Expanding the worlds AI can see

Cohere Labs

PUBLISHED
2025-03-04

SOURCE 2 · PRIMARY · CORROBORATION

Aya Vision

Cohere

SOURCE 3 · PRIMARY · CONTEXT

Aya Vision: Advancing the Frontier of Multilingual Multimodality

arXiv

PUBLISHED
2025-05-13

Trace the chronology

Record identity

LINEAiGE ID
aya-vision-launch-2025