Tech Tree Overview

Unlike a lot of 4X games with a strictly linear tech tree, Stellaris doesn’t let you “unlock tech A to reveal tech B.” Instead, every time you finish a research project, you draw a handful of random options from the relevant category’s pool — which is exactly what trips up a lot of new players.

Internally it’s a deck-and-hand system: the game holds a hidden “deck” of every technology you currently qualify for, and starting a new project deals you a “hand” of a few alternatives from it. With roughly 400 techs spread across 13 subcategories, no two campaigns deal you the same sequence — but the deck is shaped by rules you can learn and exploit.

How the three categories break down

• Physics: energy, shields, sensors, and much of the ascension-related tech lives here.
• Society: population, ethics, government, and a lot of species/bio-related tech lives here.
• Engineering: hulls, weapons/armor, buildings, and megastructure prerequisites live here.

Each category splits further into subcategories, which is where the real structure hides:

• Physics: Computing (research labs, research speed, sensors, point-defense, espionage), Field Manipulation (energy production, shields, cloaking, cosmic storm tech), and Particles (ship reactors, energy weapons, FTL).
• Society: Biology (food, genetic modification, armies), Military Theory (naval capacity, fleet command limit, claims), New Worlds (habitability, terraforming, blockers), Statecraft (unity, envoys, diplomacy), plus the exotic Psionics and Archaeostudies fields.
• Engineering: Industry (minerals, robotics, construction), Materials (armor, strategic resources), Propulsion (kinetic and explosive weapons, thrusters), and Voidcraft (ship hulls, starbases, strike craft, megastructures).

Why a key tech keeps not showing up

The tech pool is weighted-random. How many techs you’ve already researched, your policy choices, your traits, and the fixed (usually very low) weighting on certain rare techs all affect the odds of a specific tech appearing. If you’ve gone a long time without seeing the tech you want:

1. Check whether there’s a prerequisite tech you haven’t researched yet — a lot of rare techs need a prior chain finished before their weighting increases meaningfully.
2. Check whether a policy or leader trait is dragging down that category’s draw odds.
3. Consider boosting output specifically in that research category (rather than spreading evenly across all three) to increase your absolute number of draws.

A few less obvious factors sit behind those checks:

• Tier gating. Techs run from tier 0 to tier 5, and a higher tier only enters your deck after you’ve completed 6 techs of the previous tier in that same research area. No amount of luck deals you a T4 tech while your deck is still on T2.
• Anti-repeat damping. Any tech that appeared in your previous hand has its weight halved for the next draw (it resets after that), so a tech you skipped once is temporarily less likely to come back.
• Rare tech weighting. Rare techs carry roughly one-eighth normal weight by default, but several things push back: a councilor scientist with the matching Expertise trait multiplies that field’s weights (up to ×1.75 at tier III), the Technological Ascendancy Ascension Perk multiplies rare-tech weight by 1.5, and a federation’s Scientific Inspiration perk doubles the weight of rare techs another member already has.
• Permanent alternatives. Debris analysis, anomalies, special projects, archaeological sites, and situations can pin a tech below your regular hand with a golden border — it stays available until fully researched, usually with partial progress already banked. For a specific rare tech, this is often the realistic acquisition path: fight empires that field the component you want and analyze the wreckage.

Widening your hand: research alternatives

Each hand contains 3 research alternatives by default, and every +1 you stack is one more chance per hand that the tech you want shows up. The common sources: the Science Division tradition in the Discovery tree (see Traditions), the Technocracy civic, the Scientific Revolution Edicts option, the Self-Evolving Logic technology, and a handful of traits, authorities, and federation or subject arrangements. A few picks cut the other way — the Fear of the Dark origin, for example, starts you at −1 alternative.

Cost, speed, and stored research

• Cost scaling. Base tech costs are modified by the galaxy’s tech-cost setting and by empire size: every point of size above 100 adds 0.2% to research costs, so runaway sprawl slows your tech rate the same way it slows your traditions.
• Swapping projects. You can change an active research project at any time without losing progress — but progress is stored as an absolute value, not a percentage. If empire size growth inflates the tech’s cost while it sits unfinished, your completion percentage actually drops.
• Council bonus. Your Head of Research councilor adds +2% research speed per leader level (up to +20%); leaving the seat empty costs you −25% research speed empire-wide, which is one of the most punishing penalties in the game for a single empty chair.
• Stored research. Overflow output and lump-sum rewards (anomalies, debris, events) bank as stored research and pay out alongside your normal output at up to double speed until the stockpile drains — so windfalls effectively fast-forward your next few projects rather than arriving as instant unlocks.

Advice for looking up specific tech effects

Since the pool is random, it’s more useful to focus on understanding what each category of tech unlocks than to try to memorize a fixed research order — that way, whatever you draw, you can immediately judge whether it’s worth prioritizing.

To look up what a specific unlocked mechanic actually does, check the Wiki index for the matching entry.

FAQ

Is there really no fixed research order? Correct. Prerequisites and tier gating shape what’s in the deck, but within those rules each hand is a weighted random draw.

What are “dangerous” techs? A handful of late-game techs are flagged in red as dangerous — notably the advanced AI and jump-drive lines — because researching them tilts which endgame crisis the galaxy is likely to roll. They’re strong techs; just know what you’re signing up for.

Can I reroll a bad hand? Not directly, but every completed project deals a fresh hand, so finishing a cheap tech is the standard way to “redraw.” The research automation setting exploits exactly this: it always picks the cheapest available tech, unless a rare tech shows up, which it grabs instead.

Are there repeatable techs? Yes — each research area has repeatable endgame techs that keep converting research output into small permanent bonuses after the meaningful parts of your deck run dry.