Ripple 65B Preview
A 65-billion-parameter dense language model released by Recursive under the Apache 2.0 license. It supports very long context and is oriented toward programming and agentic tasks.
Timeline (last 90 days)
The Ripple 65B Preview model was developed by RicursiveAI and has officially launched in the Ollama model library.
Ripple 65B Preview is designed specifically for Agentic Coding and supports a context window of up to 1 million tokens.
Ripple 65B Preview can run locally, achieving full zero data retention.
A model of 65B parameters places considerable demands on local hardware; ordinary consumer-grade GPUs struggle to run it smoothly, so actual deployment is more likely to happen on workstations or enterprise-grade servers.
Recursive released Ripple 65B Preview, a 65-billion-parameter dense model under the Apache 2.0 license, pretrained on 12 trillion tokens and supporting up to 1 million tokens of context.
Ripple-65B Preview is a Preview version; the benchmark results come from the official release, and no independent third-party reproduction has been seen yet.
All Facts (6)
Ripple 65B Preview is designed specifically for Agentic Coding and supports a context window of up to 1 million tokens.
50%UnverifiedRipple 65B Preview can run locally, achieving full zero data retention.
50%UnverifiedA model of 65B parameters places considerable demands on local hardware; ordinary consumer-grade GPUs struggle to run it smoothly, so actual deployment is more likely to happen on workstations or enterprise-grade servers.
50%UnverifiedRipple-65B Preview is a Preview version; the benchmark results come from the official release, and no independent third-party reproduction has been seen yet.
50%UnverifiedThe Ripple 65B Preview model was developed by RicursiveAI and has officially launched in the Ollama model library.
50%UnverifiedRecursive released Ripple 65B Preview, a 65-billion-parameter dense model under the Apache 2.0 license, pretrained on 12 trillion tokens and supporting up to 1 million tokens of context.
50%