Tasks and inputs
Arrow 2 is listed for Image-text-to-image, Text-to-image. It accepts text, image and produces text, image. Use the publisher’s input format and preprocessing for your chosen deployment.
QuiverAI’s Arrow 2 generates, vectorizes, edits and animates SVG artwork through a model-specific API.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
Arrow 2 is QuiverAI’s fast model for creating and revising vector artwork. Its output is SVG, so the useful question is whether the resulting paths and document structure are editable and render correctly in your application. A thumbnail alone can hide malformed shapes, unnecessary path complexity or font substitutions.
QuiverAI lists Arrow 2 as generally available for Production and Test keys. A key must still have authority for the chosen model and operation. Check the returned supported operations and billing fields in the model API before wiring generation, vectorization, editing or animation into a product. Those operations can have different request shapes and billing conditions.
The publisher schedules Arrow 1.x retirement for 16 October 2026 at 08:00 UTC. The affected IDs include arrow-1, arrow-preview, arrow-1.1 and arrow-1.1-max. Arrow 1.0 is already rejected. Arrow 2 migration keeps the native endpoint paths, but a saved model ID still needs to change. Test existing assets and editing prompts before switching a production workflow.
Use Arrow 2 for an evaluation centered on rapid creation and iteration. Compare it with Arrow 2 Telos when complex vector construction or longer refinement context matters. Give both the same visual references, then check geometry, editability, text rendering and the number of revisions needed. The publisher’s speed positioning is not a measured latency result from SyncDev.
See the model documentation below for the current operations and access rules. Media billing is kept separate from text-token prices in this directory.
Compare the model on the work you need to complete.
Arrow 2 is listed for Image-text-to-image, Text-to-image. It accepts text, image and produces text, image. Use the publisher’s input format and preprocessing for your chosen deployment.
Documented limits: 131.07K context and 65.54K maximum output. Choose the output size and quality your design needs. Include drafts and revisions when budgeting an accepted image.
Use the model identifier and account requirements for the selected API route. Confirm its request limits and output format with one complete task before connecting a larger workflow.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
Gemini Nano Banana 2.1 is Google's image-generation and conversational-editing model released. It accepts text, images, video and PDFs and returns images and text.
QuiverAI
Arrow 2 Telos is QuiverAI’s SVG model for complex vector creation and refinement with longer context.
Qwen
Qwen Image 2.1 is a downloadable model that combines text-to-image generation, image editing and native transparent RGBA output.
OpenAI
GPT Image 2.5 Flare creates and edits images from text and image inputs through OpenAI’s image-generation APIs.
OpenAI
GPT Image 2.5 Sunburst creates and edits images from text and image inputs through OpenAI’s image-generation APIs.
xAI
Grok Imagine Image 2.0 is a hosted image-generation and editing model accessed through the Imagine API.