Ricursive releases Ripple-65B Preview, 65B dense open-weight model, scoring 80.5 on Terminal-Bench 2.1
Ricursive Intelligence released Ripple-65B Preview, a 65B dense open-weight LLM under Apache 2.0. It scores 80.5 on Terminal-Bench 2.1, ahead of US open models up to 15x its size, according to MarkTechPost.
Ricursive Intelligence has released Ripple-65B Preview, a 65B dense open-weight large language model under the Apache 2.0 license. It scores 80.5 on Terminal-Bench 2.1, ahead of US open models up to 15x its size, according to a MarkTechPost report published November 4, 2026.
Key facts
- Model: Ripple-65B Preview
- Developer: Ricursive Intelligence
- Architecture: 65B parameters, dense
- License: Apache 2.0, open-weight
- Terminal-Bench 2.1: 80.5
- Comparison: ahead of US open models up to 15x its size, per the report
Details
Ripple-65B Preview is a dense model, meaning all 65B parameters are used for each token rather than a subset as in mixture-of-experts designs. The weights are released under Apache 2.0.
The headline result is the 80.5 score on Terminal-Bench 2.1. MarkTechPost reports that this places the model ahead of US open models up to 15x larger.
What this means
The reported result pairs a 65B dense model with a Terminal-Bench 2.1 score that, according to MarkTechPost, exceeds US open models up to 15x its size. Apache 2.0 licensing means the weights can be obtained and used under that license's terms. The model is labeled a Preview, so the released checkpoint carries that designation.
Related Articles
Tencent's Hy4 adds vision and computer use, ranks first among open-weight models in blind finance evaluation
Tencent's Hy4 adds vision and computer use and is positioned in the top tier of open-weight models. On a WorkBuddy-focused evaluation it posts an average quality score of 83.18 and a 96.15% pass rate, and it ranked first among open-weight models in a blind finance evaluation by 213 experts.
Tencent releases WeVisOne-VL, Apache 2.0 vision language model in 2B and 4B sizes
Tencent's WeChat Vision team released WeVisOne-VL, an Apache 2.0 vision language model in 2B and 4B sizes. One model handles detection, OCR, counting, segmentation and alpha matting, using a 9.6M-parameter pixel decoder.
Sarvam AI releases Prism, a 105B-based decision model that claims 44 ms responses and a 71.8 average score
Sarvam AI has launched Prism, a System 1 decision model built as a LoRA fine-tune of Sarvam 105B. The company claims 44 ms response times and a 71.8 average score across three benchmarks. Weights are on Hugging Face under Apache 2.0.
Deep Principle releases Hydrogen-1.0, a 753B open-weight scientific model under the MIT License
Deep Principle has released Hydrogen-1.0, a 753B open-weight model for chemistry, materials science and life sciences. It is post-trained from Z.ai's GLM-5.3 and ships under the MIT License. The release reports a FrontierScience-Research win over GPT-6-Astra.
Comments
Loading...