Apertus 1.5 8B
The updated 8B model weights of Apertus 1.5 8B were released to Hugging Face on July 24 2026 under an Apache 2.0 license Apertus 1.5 8B matches GPT‑OSS‑20B on machine‑translation benchmarks Apertus 1.5 8B outperforms Claude Haiku 4.5 on translation benchmarks Apertus 1.5 8B outperforms Google Translate on translation benchmarks Guardian aLoRA outperforms Jev 1.13 on safety‑related decision tasks Nemotron 3.5 Super VL was developed with Mistral AI Mistral AI is a member of Nemotron Coalition· Entity summary generated by Graphiti
Relation timeline
BY VALID-FROM DATE- Guardian aLoRAOUTPERFORMSApertus 1.5 8B
Guardian aLoRA outperforms Jev 1.13 on safety‑related decision tasks
- Nemotron 3.5 Bolt 74B-A9B 24GBDEVELOPED WITHApertus 1.5 8B
Nemotron 3.5 Super VL was developed with Mistral AI
- Apertus 1.5 8BMEMBER OFNemotron Coalition
Mistral AI is a member of Nemotron Coalition
- Apertus 1.5 8BUSESNemotron 3.5 Super VL
Hebbia uses Nemotron 3.5 Super VL
- Apertus 1.5 8BPERFORMS EQUAL TOGPT-OSS-20B
Apertus 1.5 8B matches GPT‑OSS‑20B on machine‑translation benchmarks
- Apertus 1.5 8BOUTPERFORMSGoogle Translate
Apertus 1.5 8B outperforms Google Translate on translation benchmarks
- Apertus 1.5 8BOUTPERFORMSClaude Haiku 4.5
Apertus 1.5 8B outperforms Claude Haiku 4.5 on translation benchmarks
- Apertus 1.5 8BRELEASED TOHugging Face
The updated 8B model weights of Apertus 1.5 8B were released to Hugging Face on July 24 2026 under an Apache 2.0 license
Relations are facts Graphiti extracted from reports, not verbatim quotes. An end date means later reporting showed the relation no longer holds.