⚜ SteamCozy

Home / News / Stellaris

Stellaris Dev Diary #429 - Summer Experiments 2026

Stellaris · Community Announcements · August 20, 2026

Read the original on Steam ↗

by Eladrin

Hello everyone!

With summer coming to an end, the Stellaris dev diaries are back on the schedule. We’ve updated the 4.5 Open Beta a couple of times now, and have some more changes coming soon with Update 6.

Stellaris 4.5 'Cygnus' Open Beta Update 6 Release Notes

Balance

Bugfix

AI

UI/UX

Performance

Stability

Modding

The full notes so far (including the above changes) are available on the Paradox forums.

A special note of interest is that macOS builds can now run natively on Apple Silicon as a universal (arm64 + x86_64) binary, no longer requiring Rosetta 2 translation. Preliminary results thus far from the Open Beta suggest that this is an absolutely massive performance improvement for Macs, but we’d love to get more feedback.

We expect another update to the Open Beta next week. With 4.5 in active development, we’re focusing a significant amount of the Custodian designer resources on Bugfixes and AI Improvements.

Summer Experiments

[p align="center"]Everything shown below

[p align="center"]is not guaranteed to make it 

[p align="center"]into the game and even if it does, 

[p align="center"]we don’t have a timeline 

[p align="center"]at this point. 

Handing over to Alfray Stryke to fill you in on how our Summer Experiments and Custodian work is shaping up.

Hey folks, as is tradition, over the Summer the team has had the chance to experiment with ideas that may or may not make it into the game. Although these are being shown off now, as always there is no promise that they’ll make it into the game, even if features below look fairly polished, everything is a rough draft and none of this has been reviewed, playtested or verified by QA.

Harvesting Rights

A frequent request following the release of Nomads was further diplomatic improvements for Harvesting, to that end we’ve been exploring adding Harvesting Rights as a trade deal.

Under this arrangement, the Nomadic empire will be allowed to freely harvest resources within the borders of the Settled empire for the duration of the agreement. However this has some costs:

Arc Furnace Smelting Processes

Using the megastructure overclocking added for Galactic Hyperthermia in Infernals, we’ve been looking at improvements that can be made to existing megastructures and the obvious test candidate was the Arc Furnace. In this, when an Arc Furnace is upgraded to Tier 3, instead of only unlocking alloy production, it unlocks a variety of Smelting Processes, allowing tuning of the Arc Furnace production.

Smelting Processes for a fully upgraded Arc Furnace

The hope is that the code changes that enable these improvements will let both us developers and our modding community to make megastructures more interesting in future patches.

Alongside this we’ve been exploring bundling Starbase/Waystation capacity and megastructure limits into a single “Structure Capacity”, which seems fruitful, but there’s not much to show off in screenshots for this.

Mercenary Improvements

While exploring the empire flag improvements detailed below, we discovered that the flag emblem for mercenary enclaves wasn’t being used correctly. As part of the flag work, mercenary enclaves will always use their flag emblem, but inherit their flag colours from their current patron.

Following on from the auto-designer policies introduced in 4.5, we felt that allowing your to direct how your mercenary enclaves built their fleets was a natural extension.

Nanite Swarms

Back in 4.3 we reduced ship counts across the board by increasing the amount of naval capacity used for ships, with the exception of nanite swarmers These presented a problem, we didn’t want to remove the player fantasy blotting out an entire star system with your swarm, but we did want to reduce the number of ships you have. 

The solution: +500% Ship per Ship!

With some creative reimagining of how ship sections were supposed to work and some exploration by one of our animators, we’ve made a prototype in which each ship section/growth stage of the nanite swarm is visually two ships but is considered a single ship for calculations. 

Empire Randomization

Hi, pthooie here. Some players (me) like to play Stellaris with randomly generated empires, rather than making a custom design. We have a “Random” option in empire selection, but there are some problems with that. For instance, it generates the entire empire design when starting the game, so you can’t, say, choose a specific origin while randomizing everything else.

So far, the only thing you could randomize separately in empire creation was some names. Now I’ve added some new buttons:

Each of these buttons randomly chooses between all valid options for the corresponding variables. Unlike in-game randomization, this doesn’t use scripted weights, so all valid options have an equal chance of being picked.

The randomizers for ethics, civics, species traits and ruler traits randomize a full set (spending all points if possible). However, if you already have some items selected, these randomizers will not remove them. This way, you can e.g. manually pick some traits while randomizing the rest.

(Note: Layouts aren’t final; UX may move some stuff around.)

Hopefully this will help make your random empire designs more customizable!

Flag Improvements

We’ve been experimenting with making use of the flag shaders to recolor flag emblems using the existing flag colors.

NOTE: Previews below were created by using python to composite the layers and recolour them. These differ slightly from the in-game appearance.

Currently randomised flags used by AI empires are limited in the color combinations that they are allowed to use, with the allowed combinations defined many, many years ago when our color palette was much more limited (~20 colors vs. the ~70 we have now). We’re looking at broadening the allowed combinations to remove some of the restrictions, if we allow AI empires to make use of the emblem coloration above.

Flag Editor

After we showed off the above flag emblem work to our UX Designer, she felt that the flag designer in empire creation needed a bit of polish.

One thing lead to another and then Iggy came up with the idea “what if you can edit your flag in game”

This then led to conversions in the design team of “what if you could edit your subjects’ flags or the flags of those you’ve humiliated in war?”

Empire Selection and Summary

Finally from our UX Designer: Here are some experimentation/changes I have made to the empire creation UI to see if I can improve the layout to use the space more efficiently. Here is what it looks like on empire selection so far, but is still in progress: Now the UI shows the bigger City and Room view to show off your empire, with the details rearranged below.

And here is what it looks like while editing/making an empire, now you get to see a summary on the top corner as you build your empire.

I also explored collapsible categories for the many many empires we have to select from. So far it is divided between player made vs prescripted but there are more categorizations and filtrations or possible player made categories that could be done, starting from pinning similarly to Situation Log.

Espionage

Heya! Cosmogone here. Some of you might already know that I have wanted to improve our Espionage system ever since I was hired, five years ago. Over the years, every summer, I have slowly chipped at it, and this year is the year where it all coalesced into something that’s looking about ready to test. Espionage has been considered underpowered since it came out. It takes a long time to setup, and a lot of clicks to perform operations that yield lackluster rewards. The original thought process behind that choice was that it could easily become frustrating for players to get ganged up by several other empires, and so, this would be the first issue to solve to lay healthier foundations for the system.

Vigilance is meant to sort this out. It is a new value that every empire has, that tracks how alert they are to hostile spycraft in their territory. Every time an operation is completed, vigilance increases. It then slowly decays back down to its base value as things return to normal. Starting at 20 vigilance, empires start gaining passive bonuses to their encryption, the cost and difficulty of hostile operations, and at higher levels, Asset Capacity (more on this below). Vigilance usually does not increase in big leaps and ticks down fast enough that if you are the only active spy, it should not climb too fast, but as more players target the same empire, it will be harder to offset.

The other effect of vigilance is that while it is above certain values, it enables Vigilance Events to trigger. There are a variety of them, some offer bonuses to the defending country, but most will have negative consequences on active spy networks, reducing infiltration, destroying assets, taking out envoys, or outright destroying the network as a whole at the highest level of vigilance.

Vigilance should give us the opportunity to create more impactful operations, despite the risk, so naturally my next step was to buff existing operations. The list is long, but it’s things like letting Sabotage Starbase actually deal damage to starbases, giving you a precise choice of target for Smear Campaign, and enabling Asset sacrificing for extra effects in the operations that were missing them.

Next up was a general rebalance of the system. Some sources of infiltration speed that were a bit overtuned have been reduced, the base infiltration cap has been lowered, the max has been increased above 100, but the most impactful change I’d say is the broadening of the relative encryption/codebreaking gap. It previously capped out at a difference of 4, it now goes up to 6, rewarding you more for outpacing the encryption of your target. Each point rewards you with 10 Max Infiltration, 20% infiltration speed and -1 Operation difficulty, for a current max of +60 Infiltration, 120% Infiltration speed and -6 Operation difficulty. 

These massive advantages are hard to offset, even with the new vigilance system, and the game has way more sources of codebreaking than it does encryption, so naturally, I had to look at adding ways of protecting yourself. This took the form of a series of leader traits, an edict, and various modifiers spread across existing civics, traditions and ascension perks. I will not go into full details here, but there are close to 50 lines of balance changes in the changelog. A lot of work has also happened on the backend, fixing issues, edge cases, and opening the door for new things, even if I haven’t had the time to use all of it yet. For instance, operations should now be able to target leaders and colonies directly, there are new triggers and effects related to assets, some operations can be made to require an asset to start, and assets can now carry custom tooltips.

A lot of this was immediately put to nefarious use with the new Upgrade Assets. I have always wanted ways for spy networks to generate passive benefits and not force me to constantly make sure they are running operations. I’ve envisioned it in the form of “network buildings”, but ultimately fell back on the existing (and underused) Asset system. Assets are already quite powerful, raising your infiltration cap passively, and then reducing operation difficulty by up to 8 if you have the right asset on hand. The ten new Upgrade assets represent more structural upgrades to your network rather than lone operatives. They have powerful passive effects but are not meant to be used in the field, and will slowly increase their target’s vigilance. As long as there aren’t too many of them, everything should be fine, but it is definitely possible to outpace base vigilance decay, especially with several countries trying to leech off technology from a single victim. They are acquired through the new Upgrade Network operation that will sacrifice an existing asset and an investment of trade to set up the Upgrade.

But, with no limit on the amount of assets you could have, we ran into the issue of the only choice being “when” rather than “what”, so naturally, I had to implement an Asset cap. You start with a comfortable 6 assets max, but this can be further increased through civics, traditions, or the new Web Weavers ascension perk. Notably, it is also reduced by the target’s vigilance, so you might have to sacrifice valuable upgrades if the heat gets too high! This being said, the cap is technically soft and you can go over it, but will be unable to conduct any operation for as long as you exceed it.

But wait, there is more! While this covers the systemic changes, I also took this opportunity to look at some espionage QoL like:

Sub category icons for operations!

Spy Networks in the outliner, listing the target name and flag, the infiltration progress and cap, the target infiltration, and the absence of a spymaster!

Counter Intelligence in the top bar!

This one is likely to change, but there was previously no way for you to know exactly how much encryption and codebreaking you had, let alone vigilance! That’s all I have today. Hopefully, these changes will set us up to do cooler espionage things in the future, but regardless, it will all need more testing and tweaking before it is ready to play.

What’s Next?

In next week’s dev diary we’re going to continue exploring the 4.5 update.

See you then!

Stellaris MORE ON STEAMCOZYStellaris 86.0% positive · 185,433 reviews Price tracking · player charts · similar games →