Multiverse Computing has launched Quasar 438B, achieving a score of 43 on the Artificial Analysis Intelligence Index v4.1.1, the highest of any European AI model in the comparison, the company says. The new model delivers a response time of 15.3 seconds for 500 output tokens, surpassing Mistral Medium 3.5, which scores 13 points lower on the Intelligence Index. Enrique Lizaso, co-founder and CEO of Multiverse Computing, said Quasar 438B demonstrates that European AI developers can achieve both reasoning performance and speed.

Quasar 438B Achieves Top European Score on AI Index v4.1.1

Establishing it as the highest-ranking artificial intelligence model developed in Europe, Multiverse Computing’s announcement details this result. This score surpasses the 30 attained by Mistral Medium 3.5 and the 38 recorded by NVIDIA Nemotron 3 Ultra, suggesting a potential shift in the competitive landscape of AI development, which is currently dominated by US and Chinese entities.

The model’s performance demonstrates a growing European capability to produce AI systems that rival those originating from other global centers. Beyond its high Intelligence Index score, Quasar 438B delivers 500 output tokens in 15.3 seconds, a speed exceeded by only three other models in the comparison; Gemini 3.7 Flash also achieves a higher score on the Intelligence Index.

This combination of reasoning ability and processing speed is particularly valuable in applications where repeated model calls can accumulate significant delays, making even small improvements in response time crucial for efficient operation, according to Multiverse Computing. A single request in such a system may require the model to formulate a plan, utilize various tools, and refine its approach based on the results, amplifying the impact of faster processing.

Quasar 438B is designed to meet these demands, specifically targeting applications in enterprise agents and coding, and scoring 75.0 on the Artificial Analysis Long Context Reasoning benchmark, which assesses a model’s ability to reason with information from extended texts. The model supports both English and Spanish, broadening its potential reach within European markets and beyond.

Quasar 438B Demonstrates Speed and Reasoning for Enterprise Agents & Coding

Establishing it as the highest-ranking European AI model in that benchmark, Quasar 438B directly challenges the dominance of US and Chinese systems. This composite score, a weighted average of nine evaluations across four categories, indicates a broad capability beyond isolated tasks, encompassing agent performance, coding proficiency, scientific reasoning, and long-context understanding.

The model’s performance on the AA-LCR benchmark, scoring 75.0, demonstrates its ability to reason with extended documents, matching the performance of Grok 4.6 and exceeding both Nemotron 3 Ultra by 4.0 points and Mistral Medium 3.5 by 9.7 points, the company says. Beyond its overall Intelligence Index score, Quasar 438B’s speed is a key differentiator; it completes the Terminal-Bench v2.1 evaluation, assessing agent functionality in real terminal environments, with a score of 69.3.

This result places the model 18.7 points ahead of Mistral Medium 3.5 and 15.4 points ahead of Nemotron 3 Ultra, highlighting its efficiency in practical coding and command-line operations. The ability to rapidly process and execute tasks within a terminal environment is particularly valuable for enterprise applications requiring automated workflows and tool integration. Multiverse Computing emphasizes that Quasar 438B delivers both reasoning performance and speed, a combination they believe is essential for real-world deployment and broad adoption within European markets and beyond.</p, the company states.

Source: https://multiversecomputing.com/

More like this