NVIDIA has launched the Llama Nemotron Super 49B v1.5, an advanced open reasoning language model that has achieved the top position on the Artificial Analysis Intelligence Index leaderboard. This index evaluates models based on performance in advanced mathematics, science, and agentic tasks within the 70 billion parameter open model category. The Llama Nemotron Super 49B v1.5 scored 64 on the index, surpassing previous versions including the Super 49B v1 released earlier in the year. NVIDIA has also made over 26 million lines of synthetic data used in training the model publicly available, enhancing transparency and enabling developers to build their own models more efficiently. The model is commercially ready and accessible via Hugging Face, featuring best-in-class performance for reasoning and agentic tasks. Additionally, updates have been made to related open-source repositories to support replication of results on open models.