Changelog
Follow the latest features, improvements, and fixes as Novelcrafter evolves.
Today’s update refreshes the system prompt model lists with the following additions and changes:
Removed models that are no longer supported by their vendors (e.g. older Gemini and Mistral).
Included more modern and alternative models that perform well at their tasks (e.g. GLM, MiMo and Aion)
Existing Chats using removed models from the old system collection will keep their settings.
As always, you can find previous system model collections in the archive.
Replaced the “Try for a spin” with a search modal for easier access than a menu. This also fixes the issue of the site hanging when listing models from Featherless.
You can now pick any one-off model for a Chat thread, without the need for a model collection.
Fixed the pov prompt functions not working with equality and contains checks - especially when comparing POVs with characters. This should improve the output of the default scene beat prompt when dealing with multi-POV stories.
Disabled auto-cache-write requests for GPT 5.6/6/6.1 models when used via OpenRouter. Previously, cache writes were offered for free, but now they’d cost 25% extra.
Updated the AI vendor icons for OpenRouter, Google AI Studio and Featherless, since they’ve all changed their branding.
It’s been quite some time since our last big update. Even though most of our changes with this one are kind of behind-the-scenes, we do have one major thing that’s new:
Today’s update overhauls how you create, move, and organize your acts, chapters and scenes!
Read the full guideDon’t worry - it largely looks and feels the same. However: you now have way more options to organize your book from either, the Plan or the Write view. Who doesn’t like less back-and-forth hopping?
Today, we’re enabling two long-awaited experimental Codex features for everyone: better control over tracking your entries, as well as allowing an entry to be in more than one custom category at a time!
Kate will also go into detail on both of these updates in a live stream later today.
For those that wanted to have a “Don”, “May” or “Will” as part of their stories, have probably quickly realized that Novelcrafter picked up on a lot more mentions than it should have. For these kinds of entries, we added two new options:
So, if you’ve got a “Don”; simply enable case-sensitivity, add “don’t” to the excluded phrases, and you’re good to go!
You can learn more about these options in our updated and expanded documentation.
As always, massive thanks to our testers for finding various edge cases and weird names to test against!
Sometimes, a character or location just doesn’t fit in a singular category. They may be part of a faction as well as belong to a group of friends. Fear not! Novelcrafter will now simply put them in all the categories that match.
There’s no options here, this has now been enabled by default.
We just shipped an update that adds support for bullets and numbered lists in the Word / .docx exports! Feel free to try it out with your document and report any bugs or if we missed something.
Here’s more that we shipped recently:
Today, we’re releasing our experimental support for thinking/reasoning to everyone. As all experimental features, you can enable it in your Settings.
Since models differ a lot in how they do their reasoning, we chose a more general approach: Once opted in, you’ll see an option to ‘prefer’ or ‘avoid’ reasoning across your Scene Beats, Chats, and other areas in Novelcrafter.
To tweak the reasoning level, you can do so via the custom model parameters (just like changing temperature).
For more details, please check out our new documentation here.
As always, many thanks for our early testers for their feedback and bug reports!
We just updated the system model collection to include the new Gemini 3 Pro model. The update also includes an update of GPT-5 to GPT-5.1 (where applicable).
As always, if you want to revert to the previous collections, read about how to do so on our FAQ article.
Today we updated the default AI model collections. The new collections now include Claude 4.5 Sonnet, contain updates to newer versions (e.g. the new Gemini 2.5 Flash) and have been cleaned up to remove models with very low usage.
Remember that you can create your own (we recently published a video about how to do this on YouTube), and that your own collections are unaffected.
If you want to revert to the previous collections, read about how to do so on our FAQ article.
In case you missed it, we also posted a community update blog post recently.
We haven’t posted an update in a while, so here’s what’s new today (plus some fixes we’ve made over the past few weeks):
We’ve updated and enhanced the inner workings of all AI connections:
[object Object].While we continue to enhance our prompting system (and its documentation), we’ve cleaned up a few things and added/fixed some functions:
Here’s what’s new this week: