See what's new in FLORA ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏
|
| | Exciting updates now live in your FLORA canvas. |
|
| | | | | This past month we added a wave of models that help expand the possibilities in your creative workflows. |
|
| | Nano Banana Pro brings Google's Gemini 3 Pro reasoning to image generation. Beyond the aspect ratio control we added earlier, Pro delivers studio-quality text rendering in multiple languages, connects to real-world knowledge for accurate infographics and diagrams, and lets you upload up to 14 reference images to maintain brand consistency. Think of it as few-shot prompting for designers—load your style guide, logos, and product shots, and the model maintains that identity across outputs. Additionally, it can achieve a perfect hex code extraction from any image, making it a milestone for diffusion models in general.
This is a big leap forward. We'd love to hear your feedback.
|
|
| | | | Seedream 4.5 is ByteDance's latest commercial-grade image model, purpose-built for professional creators. It excels at text rendering—sharp, legible type for posters, logos, and marketing layouts without Photoshop cleanup. The real standout is multi-frame consistency: generate multiple images from a single prompt while keeping characters, clothing, and environments stable across the batch. Up to 4K output and 30% faster inference than its predecessor.
Flux 2 from Black Forest Labs landed with full multi-modal support: text-to-image, image-to-image, and image+sketch-to-image in both Dev and Pro tiers. Dev mode now accepts up to 3 inputs for more complex compositional work.
Qwen Edit Angles gives you precise camera viewpoint control on existing images. Adjust perspective with a numerical slider without regenerating from scratch—ideal for product shots and architectural visualization. |
|
| | Gemini 3 Pro is Google's most capable reasoning model yet, with deeper contextual understanding and stronger instruction-following.
Claude Opus 4.5 is Anthropic's flagship—their most intelligent model with exceptional nuance, creativity, and extended thinking capabilities. When you need careful reasoning through complex creative briefs or multi-step workflow logic, Opus 4.5 brings a level of depth that shows in the output quality. We'll have more on this next week.
Claude Sonnet 4.5 balances strong reasoning with speed, making it ideal for rapid iteration. |
|
| | Kling O1 is Kuaishou's latest family, spanning text-to-video, image-to-image, image-to-video, and edit modes. Noticeably better motion coherence and prompt understanding than the 2.5 generation—particularly on complex scenes and longer outputs. |
|
| FLORA 2.0: Built for how you actually work |
|
| We've been shipping foundational upgrades that make the canvas faster, smarter, and more collaborative.
Private sharing that works like docs. Share projects via editable links without making them public. Invite specific collaborators—including people outside your workspace—while keeping your work hidden from everyone else. Or, share with anyone! True selective access. |
|
| | In-block generation history. Every block now stores up to 100 past generations. Scroll through them in a carousel, navigate with keyboard shortcuts or trackpad gestures, and duplicate or export any result from the new menu. |
|
| | Keyboard shortcuts. Add blocks, group selections, duplicate, delete, rename, fullscreen preview, upload, remove background—all from the keyboard. Hit the Help menu (or use command + /) for the full reference. |
|
| | Drag-to-replace. Drop a local file onto any image or video block to swap its asset. No rewiring, no rebuilding—just drag and update.
Smarter image loading. Images now load at the resolution your zoom level actually needs. The canvas stays responsive with high-res assets, preloading the next tier as you zoom in.
Cleaner navigation. Breadcrumbs in the header, pagination across workspace views, and automatic workspace switching when you accept an invite. Small touches that add up. |
|
| Join us at FLORA Training Sessions |
|
| | We're back with more FLORA Training Sessions—focused sessions where our team demonstrates FLORA techniques, you practice with hands-on activities, and we help answer your workflow questions in real-time.
Sessions take place weekly on Fridays → See the full schedule on Luma |
|
| | | | | | |
|
|
|
| |
|
Tidak ada komentar:
Posting Komentar