QuiverAI is building the future of vector design with AI-powered tools for generating, editing, and animating editable SVG assets. Convert text prompts into production-ready vector graphics including logos, illustrations, typography, and animations. Also supports image-to-SVG conversion.
Free tier includes 20 SVGs per week, with paid plans starting at $20/month for higher limits and additional export formats. Backed by a16z with $8.3M in seed funding.
Pricing model: Freemium
Common questions about QuiverAI
QuiverAI’s API Platform now shows a Model training row in Data Controls, replacing Inference data. The Improve QuiverAI setting controls whether the company may train on your organization’s inference inputs and outputs.
It’s on by default. An organization Owner or Admin can turn it off unless the organization is blocked. Turning it off stops training on that data, but storage still follows your other data settings. Preventing customer content from being stored requires a contract.…
If your client reads QuiverAI's data posture response headers, switch their prefix from x-quiver- to x-quiverai-. This affects inference-data-policy, payload-capture, request-metadata-policy and request-metadata-retention-seconds. Sandbox responses use x-quiverai-environment: test.
The old names continue to carry identical values until December 31, 2026, when they're scheduled for removal.
The x-quiver-zero-data-retention header has already been removed.…
QuiverAI has added Arrow 2 Telos for demanding vector designs. The company says its refinement layer handles complex briefs, detailed compositions, and larger inputs, with a 1 million token context window.
QuiverAI recommends base Arrow 2 for speed and high-volume work. Both support SVG generation, vectorization, editing, and animation.
Read the original announcement →
QuiverAI scheduled Arrow 1.x API retirement for 08:00 UTC (10:00 Berlin) on October 16, 2026. The affected IDs are arrow-1, arrow-preview, arrow-1.1 and arrow-1.1-max. At the cutoff, they stop serving, disappear from GET /v1/models and return 404 modelnotfound. Requests aren't automatically routed to Arrow 2.
The replacement is arrow-2 for SVG generation, vectorization, editing and animation. Native endpoint paths stay the same, but Arrow 2 uses measured token pricing instead of Arrow 1.x's…
QuiverAI's API now has a POST /v1/responses endpoint compatible with stock Open Responses clients. You can declare function and custom tools, then execute functioncall items yourself. The endpoint is stateless: previousresponseid and store: true are rejected.
Arrow 2 and Arrow 2 Telos use measured token billing rather than fixed credits, with rates available as billing.rates on GET /v1/models.
Sandbox keys let you test the same routes without reaching a model.…
QuiverAI's App now starts with a conversation. Describe what you want, watch the drawing stream in, then refine the editable SVG with another prompt or open it on the canvas.
On the canvas, you can select and duplicate elements, undo edits, and animate and replay a creation. Saved edits flow back into the chat and creation sidebar.
Arrow 2 is now the App's model. Arrow 1, Arrow 1.1 and Arrow 1.1 Max are retired from the App but remain available through the API.…
QuiverAI has released Arrow 2 for generating and editing SVGs, alongside Arrow 2 Telos for more demanding styles and compositions. QuiverAI says Arrow 2 generates editable vectors faster, with cleaner geometry.
The updated app brings generation, conversational iteration, and canvas editing into one workspace. Developers also get a new API platform for managing credentials, usage, and billing.
Read the original announcement →
QuiverAI's Arrow 1.1 cuts text-to-SVG generation costs by 33.3% and vectorization costs by 50%.
The model uses more shapes and text rather than paths. QuiverAI says that makes the SVGs easier to edit, inspect, and extend, with improvements to prompt following, layout balance, and color usage.
There's also Arrow 1.1 Max for detailed assets such as technical diagrams and illustrations. QuiverAI says it offers more accurate color, tighter alignment, and fewer artifacts, at a higher cost and…
Other products to explore and follow.
