π Big news: @simtheoryai and @ortto are joining Canva. Together, they bring agentic AI and full-cycle marketing automation to our platform, and there's more to come at Canva Create on April 16. Read more: https://t.co/dUukEVnsmW
We've been acquired by Canva.
Simtheory isn't going anywhere. Simtheory will continue as an independent product.
AI has moved from novelty to genuine productivity. Simtheory is your agentic AI workspace - every model, every tool, and agentic workflows that allow you to collaborate and delegate real work.
Workspace can be customised for your team and organization and lead to AI transformation within your business. Branded, secure, run on your own cloud servers with your own models.
Try Simtheory at: https://t.co/WUVvIPz8jP
Kimi K2.5 is now available in your AI workspace. It's comparable to major USA lab models and has 256k context window with 32k output tokens.
It's fast, smart and great at agentic tool calling.
Your AI workspace has been upgraded:
- GPT-5.1 and GPT-5.1-thinking now available
- Shopify, Hubspot, Docusign, Lemonado MCPs
- Scholar AI for searching 250M+ peer-reviewed articles, patents, and book chapters
- Session finder MCP to bring into context all previous sessions
- Theme maker MCP to theme your own workspace
Claude Haiku 4.5 is now available in your AI workspace.
Similar to Sonnet 4.5 it excels at agentic tasks and tool calling. It punches above it's weight in benchmarks (it's said to be comparable to Gemini 2.5 Pro at agentic coding) and has incredibly fast output.
Best of all it's marked as non-frontier meaning you can use it with no impact on your token limit.
Claude Sonnet 4.5 is now available in your AI workspace.
It's said to be the best coding model available today and excels at longer running agentic tasks including document analysis and research.
A max thinking "tune" is also available.
Try it now: https://t.co/d9PjsROSG6
A recap of updates to your AI workspace this week:
- You can now train, save and manage your own voices for use in Audiobooks, Podcasts and Videos
- We've consolidated images models to *Image tool* which will automatically select the best model for the task
- We've added Perplexity's new search API and upgraded Grok DeepSearch to use Grok FAST 4
- We improved model response rates on average by 5x
- Google's Gemini Flash 2.5 Preview is now in the model catalogue along with Grok 4 FAST models
- Kling 2.5 Pro and Omnihuman video models are now available