GTA 6 review scores: nothing exists yet, and here is what will be published

7 min read·Updated 2026-11-15

GTA 6 launches on Thursday 19 November 2026, and the review embargo lifts the day before. As of 15 November, no outlet anywhere has published a score for the game. This page exists to be useful anyway: it explains what will appear on 18 November, what reviewers are actually testing, and why the two big aggregators will probably give you two different numbers.

Status, 15 November 2026: pre-embargo. The embargo lifts on Wednesday 18 November 2026, one day before release. Rockstar has not published an hour, and the reports covering the date give none, so the exact minute is unknown. What can be said precisely is the shape of that day: 18 November in New York (EST, UTC−5) runs from 05:00 UTC on the 18th to 05:00 UTC on the 19th, and 18 November in Los Angeles (PST, UTC−8) runs from 08:00 UTC on the 18th to 08:00 UTC on the 19th. A score cannot legitimately appear before the embargo lifts, so anything carrying a GTA 6 number today is fabricated, including numbers that look precise.

What is confirmed before the first score exists

None of the following is a judgement about quality. It is the set of facts a reviewer will be reacting to, and the background you need to understand why the scores land where they land.

Every mainline GTA launch score, by platform

This is the part of the page that has numbers today. Metacritic scores each platform version separately, and the counts differ because each version was reviewed by a different set of outlets. The figures below were read from the individual Metacritic platform pages on 15 November 2026.

GamePlatformMetascoreCritic reviews
Grand Theft Auto IV (2008)PlayStation 39864
Grand Theft Auto IV (2008)Xbox 3609886
Grand Theft Auto IV (2008)PC9040
Grand Theft Auto V (2013)PlayStation 39750
Grand Theft Auto V (2013)Xbox 3609758
Grand Theft Auto V (2013)PC9657
Grand Theft Auto: San Andreas (2004)PlayStation 29580
Grand Theft Auto: San Andreas (2004)Xbox9358
Grand Theft Auto: San Andreas (2004)PC9347

These numbers are a historical record of three older games reviewed by an older press, and they are not a forecast for GTA 6. The counts move over time as Metacritic adds or removes outlets, so a page checked a year from now may not match this one. The pattern worth noticing is narrower: the PC version has scored lower than the console versions in all three cases, by one point for GTA V, by eight for GTA IV, and by two for San Andreas. GTA 6 has no PC version at launch, which removes that particular risk from the first round of reviews entirely.

What reviewers will actually be testing

"Is it good" is not a reviewable question at this scale. These are the specific, checkable things that will separate a 90 from an 85, and they are what to read for instead of the number at the bottom of the page.

  1. Frame pacing, not frame rate. A locked 30 with even delivery is playable; a 30 that drops in traffic, weather and crowded interiors is not. Look for whether the reviewer names the scenes where it fell apart, or just repeats the target.
  2. Whether the map is filled or merely large. Twice the size of GTA V is easy to advertise and hard to populate. The question is how much of Leonida carries missions, interiors and reasons to stop, and how much is scenery you drive past once.
  3. Whether the Criminal Profile and the relationship scale change anything observable. Both are confirmed to affect how the story unfolds, but a system that alters a few lines of dialogue is a different product from one that changes missions or outcomes. This is the single biggest unknown going into review week.
  4. How many approaches a mission actually accepts. Open-world games advertise freedom and then script one door. Reviewers who play a heist twice and report what changed are telling you more than any score.
  5. The state of the launch build. Every Rockstar release arrives with bugs; the question is whether they break a mission or a frame. Load times on a 212 GB install, and stability across a long session, belong in this category.

How a Metascore is built, and why OpenCritic will disagree

Metacritic describes the Metascore as a weighted average of reviews from selected critics and publications. Each review is converted to a 0–100 scale, and publications are given different weights based on how Metacritic rates their quality and standing. Those weights are not published, and Metacritic has declined to disclose them. Two details matter for games specifically. First, every game score on the site comes from the publication itself: unlike film and music, Metacritic does not invent a number for an unscored game review, which is why unscored reviews sit on the page without entering the average. Second, the normalisation that Metacritic applies — the "graded on a curve" adjustment — is described for its film and music sections, not for games.

OpenCritic works differently by design. Its headline figure is the Top Critic Average: a simple average of the numeric scores from the outlets it designates as top critics, converted to a 100-point scale, with no publication weighted above another. It also publishes a separate percentage of critics who recommend the game, drawn from a wider pool that includes outlets outside the top-critic list and some unscored verdicts. The practical consequence is that the same set of reviews can produce two different headline numbers, and neither is wrong — they are answering slightly different questions with different populations. When both go live on 18 November, a gap of several points between them will be arithmetic, not controversy.

The method pages are worth reading once: Metacritic's explanation of how Metascores are computed, and a side-by-side breakdown of how the two aggregators differ.

What this page will publish on 18 November

The rule is simple and it will not bend for a fuller-looking table. A score appears here only if someone on this site has opened the original review or the outlet's aggregator entry and read the number there. No second-hand scores, no "reportedly", no ranges presented as results. Each row will carry a link to the source and a timestamp, the table will be sorted by score, and if only six outlets have published at the moment this page is updated, the table will have six rows.

What will not appear, today or on 18 November: a predicted score, a predicted range, or any sentence beginning "critics are expected to". Pre-order records and trailer view counts are evidence of marketing, not quality, and the most pre-ordered game in PlayStation Store history can still disappoint. The embargo date itself is the only scheduling fact here, and it has been reported consistently since it surfaced in October.

If you are deciding what to do with the next four days rather than what to read on Wednesday, the practical pages are preload and storage, release time by region, and PS5 vs Xbox Series X. If your question is whether there will be anything to play with friends, that is GTA 6 Online, and the answer has not changed.

Common questions

When do GTA 6 reviews go live?

The embargo lifts on Wednesday 18 November 2026, one day before the 19 November launch. Rockstar has not published an hour, and no outlet has reported one.

What is the highest Metascore a GTA game has received?

Grand Theft Auto IV holds it, at 98 on both PlayStation 3 and Xbox 360, based on 64 and 86 critic reviews respectively as of 15 November 2026.

How do Metacritic and OpenCritic end up with different scores?

Metacritic uses a weighted average with undisclosed publication weights, while OpenCritic averages its top critics without weighting. Different methods on a similar review pool produce different numbers.

Does GTA 6 have an online mode at launch?

No. Rockstar describes the 19 November release as a single-player experience, and no date has been announced for a GTA 6 online mode.

Does this page publish a predicted score before the embargo?

No. Only scores read at the original source are published, and any GTA 6 number circulating before 18 November is fabricated.