Ripple-65B Preview delivers strong agentic coding results as open US model
Ricursive Intelligence says Ripple-65B Preview scored 80.52 on TerminalBench 2.1, was pretrained on 12 trillion tokens, and is available on Hugging Face under Apache 2.0. Ricursive says total development and training costs were $32 million.
AI chip design firm Ricursive Intelligence has trained an open language model called Ripple-65B Preview. The company says the model achieved a score of 80.52 on the TerminalBench 2.1 coding benchmark, the best result among open US models. Ripple-65B Preview is available on Hugging Face under the Apache 2.0 license.
Training costs and harness
Ripple-65B Preview was pretrained on 12 trillion tokens. According to Ricursive, total development and training costs amounted to $32 million, covering all research and development expenses. Ripple partly uses Terminus 2 and partly OpenCode as its harness.
Benchmark comparison
Other benchmark results show that Ripple does not match the also newly released Inkling 2 Small, but Ripple is likely easier to deploy on users' own hardware. On SWE-Bench Pro v2.0 Hard, Ripple scored 79.1% and Inkling 2 Small scored 86.3%. On DeepSWE v1.1, Ripple scored 44.3% and Inkling 2 Small scored 72.1%. On MCP Atlas, Ripple scored 71.67% and Inkling 2 Small scored 85.4/85.0%.
Architecture and availability
The model uses a dense Transformer architecture with 65 billion parameters instead of the more common mixture-of-experts approach. It is optimized for Nvidia hardware. Ripple-65B Preview is available on Hugging Face under the Apache 2.0 license.
Sources: The Decoder
SiTech — AI-powered web development
We build fast, modern websites and bring AI into real business workflows. Have a project or a question? We'd love to help.