ZMIME

Hy4 vs Ripple-65B Preview vs SWE-2 Flash

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

Model specifications
SpecificationHy4TencentRipple-65B PreviewRicursive IntelligenceSWE-2 FlashCognition
Task—Language generation and codingSoftware engineering
Published capabilitiesA next-generation multimodal productivity model with significantly enhanced Agent and complex task execution capabilities.Dense transformer trained on 12 trillion tokens with a focus on long-context software engineering.Coding agent model developed from Inkling-Small, with mid-training on scientific and software engineering data.
Model size—65B parameters—
Access—Public weights and hosted inference via Together AI, OpenRouter and FireworksDevin Desktop and CLI
Sources—Publisher documentation · Model weightsPublisher documentation
InputsText, ImagesTextText
Context window1,048,576 tokens1,048,576 tokens—
Input price per million tokens$0.834——
Output price per million tokens$2.50——
WeightsProprietaryOpen weightsProprietary
ReleasedNov 4, 2026Nov 4, 2026Nov 4, 2026