# Lovart updates Every Lovart update in one feed — release notes, features and changes, straight from the source. 20 updates - ## [Lovart adds Kling 4.0 with 30-second clips and 15 references]() Kling 4.0 is live on Lovart, with single-take clips up to 30 seconds, 10 keyframes, up to 15 references, and Arti drafting the shot list. 2026-10-24 - ## [Lovart adds Delegated Tasks so Arti runs recurring design work]() With Delegated Tasks in Lovart 2.0, you set up a recurring design task once and the Arti design agent runs it on your chosen schedule. 2026-10-23 - ## [Lovart’s Arti agent now turns repeated workflows into Skills]() Self-evolving Skills in Lovart 2.0 let the Arti design agent save the workflows you repeat as reusable Skills you can run again anytime. 2026-10-22 - ## [Lovart’s plugin brings brand visuals and video generation to Codex]() The Lovart Plugin is live in OpenAI’s Codex, so you can generate brand visuals, ad sets and videos where you code. 2026-10-22 - ## [Lovart 2.0 introduces Arti, a personal design agent]() Arti learns your taste and how you work, and you can schedule a brief before logging off to wake up to research and drafts ready to review. 2026-10-19 - ## [Lovart MCP Is Live on Codex]() Lovart's Codex integration lets creators generate and evaluate branded assets in batches, powered by Jev. 2026-09-23 - ## [ChatGPT Images 2.5 on Lovart]() Lovart adds ChatGPT Images 2.5, with tools for sketch-to-scene conversion, vector scaling and precise, quality-preserving edits. 2026-09-10 - ## [Lovart adds Seedance 2.5 for native 30-second 4K video]() The design agent's Seedance 2.5 integration accepts up to 50 reference assets and renders 4K natively, unlike Runway's same-week 720p version. 2026-08-07 - ## [Lovart adds MiniMax H3 for cheaper 15-second cinematic video]() The model renders up to 15 seconds with multimodal references and consistent characters, at less than half the usual cost. 2026-07-30 - ## [Nano Banana 2 Lite lands in Lovart]() The model generates images in about four seconds inside Lovart, per an X post on 2 July 2026. 2026-07-02 - ## [Lovart adds Gemini Omni Flash video generation]() Users can now turn Lovart designs into short AI video clips with conversational, natural-language editing. 2026-07-01 - ## [Lovart adds Mockup, placing designs onto real-world objects]() The new tool skips template files and Photoshop compositing, going from prompt to preview on the object itself. 2026-06-25 - ## [Lovart adds a cheaper Seedance 2.0 Mini for batch runs]() The Mini variant is built for prototyping full series and ad campaigns before a final pass through the full Seedance 2.0 model. 2026-06-23 - ## [Lovart to add Google's Gemini Omni Flash]() The AI design tool teased the integration on X with no timing or feature details yet. 2026-06-21 - ## [Lovart lets creators save chats as reusable skills]() Create Skill turns a successful conversation into a repeatable workflow that can be run again on new projects. 2026-06-06 - ## [Lovart adds Kling 3.0's native 4K video model]() The model holds subject, style, colour and mood steady at 4K for large-screen and professional work. 2026-04-27 - ## [Lovart adds GPT Image 2 with unlimited 2K generations]() The upgrade improves text rendering and texture realism, useful for logos and social assets on Lovart's Standard tier. 2026-04-21 - ## [Seedance 2.0 in Lovart now renders native 1080p video]() The update sharpens edges, removes texture flicker, and clarifies motion for professional video output. 2026-04-18 - ## [Lovart adds one-click custom font generation]() Fonts can be generated from images, text, or both, and save automatically to the user's My Fonts library. 2026-04-18 - ## [Lovart adds PSD export with editable layers]() AI-generated logos, illustrations, and social posts can now be opened in Photoshop with each element on its own layer. 2026-04-15 [Open Lovart updates on Creators Toolbox]() --- Canonical source: https://creatorstoolbox.com/updates/lovart