ZMIME

SWE-2 Flash vs Ripple-65B Preview

Compare documented capabilities and access. A dash means a comparable figure is unavailable.

Model specifications
SpecificationSWE-2 FlashCognitionRipple-65B PreviewRicursive Intelligence
TaskSoftware engineeringLanguage generation and coding
Published capabilitiesCoding agent model developed from Inkling-Small, with mid-training on scientific and software engineering data.Dense transformer trained on 12 trillion tokens with a focus on long-context software engineering.
Model size—65B parameters
AccessDevin Desktop and CLIPublic weights and hosted inference via Together AI, OpenRouter and Fireworks
SourcesPublisher documentationPublisher documentation · Model weights
InputsTextText
Context window—1,048,576 tokens
Input price per million tokens——
Output price per million tokens——
WeightsProprietaryOpen weights
ReleasedNov 4, 2026Nov 4, 2026