Framer Agents now use far fewer credits on larger, multi-step builds, giving you more room to create pages, make site-wide updates, and refine your work within the same credit budget. We worked directly with OpenAI to improve model efficiency in Framer, resulting in OpenAI models using around 25% fewer credits on multi-step builds. Alongside that work, we improved all models to complete tasks in 15–30% fewer steps, use 30–40% fewer tokens, and save another 10–15% through better caching. Agents now also catch and resolve more unintended side effects.

We’ve also made GPT-5.6 Terra the default model in Framer. In our testing, it is more capable than GPT-5.5, twice as fast, and uses 37.5% fewer credits. Combined with the broader efficiency improvements, multi-step builds with Terra use 60–67% fewer credits than GPT-5.5 did when Framer Agents launched. Sonnet 5 and Opus 4.8 benefit from the same improvements, using 48% fewer credits. These improvements are now live across Framer Agents and External Agents.

We’re introducing mouse interaction to two more Shaders: Liquid Gradient and Logo Gradient. Liquid Gradient creates animated color fields that move across your screen, while the Logo Gradient adds shifting gradient effects to SVGs and PNGs. Both respond to your cursor with smooth, spring-based tracking, combining motion-style interaction with WebGL effects. Tweak the colors, motion, intensity, and more to create endless variations of these shaders. To add the new interactions, simply click + on the new Mouse property while having your shader selected.

Adding new CMS items is now faster. A new empty row appears at the bottom of every collection, allowing you to create items directly from the table. To quickly add an item, focus the empty row, fill in one or more fields and hit Return to quickly add it and move to the next one. It’s especially useful for rapid creation of simple entries like categories, labels or tags. We’ve also shipped a number of additions, improvements, and fixes, with a particular focus on the Agent experience in the CMS. Additionally, we made formatted text fields behave more reliably. See the full list below.

Added

• Added a bottom empty row to quickly create new items

• Added Agent support for adding CMS rows as context

• Added Agent support for working with field dividers

• Added Agent support for collections via Select tool

• Added Agent support for previous and next pages

• Added support to delete multiple items with ⌫

• Added tooltips to the formatted text toolbar

• Added ability to force delete fields in use

• Added support for searching collections

• Added support for filtering by date field

Improved

• Improved Agent single-cell selection and editing

• Improved Agent duplicating synced collections

• Improved Agent awareness of selected rows

• Improved Agent awareness of draft statuses

• Improved Agent highlighting status changes

• Improved Agent handling of required slugs

• Improved Agent selected row highlights

• Improved Agent Select tool behaviour

• Improved selection of links in rich text

• Improved contrast in the item overlay

• Improved the / menu in rich text

• Improved the menu to filter items

Fixed

• Fixed inline editing of formatted text with components

• Fixed an issue with CSV exports and <video> tags

• Fixed undo when pasting images in formatted text

• Fixed a font issue when inline editing code blocks

• Fixed undo after adding a link in formatted text

• Fixed scrolling when reselecting cell groups

• Fixed an issue with scrollbars in embeds

• Fixed the Fields header not being sticky

• Fixed Add to Agent button behaviour

OpenAI’s next generation of models is now available in Framer: GPT 5.6 Sol, GPT 5.6 Terra, and GPT 5.6 Luna. Together, they give you more control over how you spend credits, whether you need the highest-quality creative output, balanced intelligence at lower cost, or maximum speed for large updates. GPT 5.6 Sol is our strongest creative model yet. In testing, it produced refined designs with minimal guidance and became the first model to score 100% on our hardest internal benchmark, a major step up from GPT 5.5 at the same credit cost. GPT 5.6 Terra is built for balance. It delivers GPT 5.5-level results while using only 0.6× the credits, making it ideal for large audits, redesigns, and consistency passes. GPT 5.6 Luna is our fastest model in Framer, running 2× faster than any previous option while using only 0.4× the credits compared to GPT 5.5. It is especially well suited for large CMS updates and complex cross-site find-and-replace tasks.