← Back to articles
Add to favorites

OVHcloud unveils Olala mini, its first AI model, already used by Qwant

OVHcloud unveiled Olala mini 7A1B, a language model with seven billion parameters, already used by Qwant for its summarised search answers.

By GennnAi··1 min read

According to OVHcloud, switching to Olala mini cut Qwant's waiting time by 20% and divided costs by three compared with the previous model. The hosting company is preparing a version of at least 100 billion parameters, focused on business agents, cloud DevOps and cybersecurity. The model is released under an Apache 2.0 licence, according to other outlets, and OVHcloud announces no pricing.

Source: Journal du Geek, published on November 19, 2026.