Research Explainer · XiaoHu Explains

Same models, new shell: Opus 4.8 and Fable 5 hit 99% on ARC-AGI-3

A harness called Schema makes the model write each game's rules as a runnable program, verify them, then act. Self-reported 98.98% on the 25 public games — not independently verified by ARC Prize.
60-second summary
  • ARC-AGI-3 is a class of games that tell you nothing: the AI only gets a 64×64 colored grid and a few buttons it can press — no object list, no stated goal, no reward signal. When it launched in March, the strongest models scored only 0.51.
  • Schema is a harness — it doesn't touch the model's weights, only how the model is used. It makes the model write each game's rules as a runnable program, replay every recorded step against that program to check it, then search within that program for the next move.
  • On the 25 public games, Opus 4.8 paired with Fable 5 scored 98.98, and GPT-5.6 Sol scored 95.35. Both are the team's own self-reported numbers, not independently verified by ARC Prize.
  • With the same Opus 4.8 and Fable 5, swapping in the stock Claude Code harness scores only 42.83. Swapping in Schema alone adds 56.15 points.
  • That 99% measures step efficiency — the AI's step count compared to a human's first-time playthrough (squared ratio). It means nearly every game was cleared at an efficiency matching or beating humans; solving it alone isn't enough.
This is Impossible Research's own harness release. Both scores are self-reported by the team, measured only on the 25 public games, and have not been independently verified by ARC Prize. How it would score on the semi-private set hasn't been measured yet.
Impossible Research × Berkeley × CMU

A game that won't tell you the rules

Impossible Research (working with UC Berkeley and Carnegie Mellon University) released Schema, a harness that lets frontier models play games like physicists. It targets a benchmark built specifically to trip up AI: ARC-AGI-3.

ARC-AGI-3 throws a game at the AI without explaining what it's looking at. Each step, all the AI gets is a 64×64 grid (16 colors) and a list of which buttons it can currently press. No object list, no rules, no stated goal, no signal telling it "that step was right." The only way forward is to act like a physicist: poke at it, guess the rules, and revise when you're wrong.

The image below is exactly what the AI actually sees across the 25 public games. Each tile is one game, and the AI has to work out for itself: which of these colored blocks is the character it controls, which are walls, which bar is a countdown, and what even counts as clearing the level.

Mosaic of the 25 ARC-AGI-3 public game screens
The 25 public ARC-AGI-3 games, one tile per game — this grid is all the AI ever sees, with no text explanation attached. Source: Impossible Research / Schema

This benchmark is brutal on frontier models. Its official metric, RHAE (Relative Human Action Efficiency), compares the AI's step count per level against a human playing for the first time. When it launched in March, the best verified frontier model scored only 0.51; by July, the best model — GPT-5.6 Sol at its highest reasoning setting — reached just 7.78 on the semi-private set and 13.33 on the public set. For reference, human action efficiency is a perfect 100.

Schema moved that number into an entirely different range: on the 25 public games, Opus 4.8 paired with Fable 5 scored 98.98, and GPT-5.6 Sol scored 95.35 — without changing a single byte of the underlying model weights.
🔧

What changed isn't the model — it's how the model gets used. Schema is a harness: it restructures how the visuals get fed in, how the model turns observations into an understanding of the game, how that understanding gets checked against real records, and how the plan built from it gets executed and corrected. Swap out this harness on the exact same Opus 4.8 plus Fable 5 pairing, and the public score jumps from 42.83 to 98.98.

Official Schema demo (47 seconds): the AI plays while writing the rules as a program — on screen it moves a block through a maze, the yellow bar at the bottom is the movement budget; the "99 RHAE" in the bottom right is the public score for Opus 4.8 + Fable 5, next to Sol's 13%. Source: Impossible Research / Haiwen Feng (@HavenFeng)
Understanding the 99%

What that 99 actually means

It's not "how many of the 25 games got solved." RHAE measures step efficiency, with a heavy squared penalty built in.

Here's the formula: for each level, take "human step count ÷ AI step count," square it, capped at 1.15. Later levels within a game carry more weight (increasing from 1 to n), and there's a completion cap tied to that weight — miss even one level, and the whole game can't score full marks. Every action the AI actually takes in the game, including exploratory steps taken just to probe the environment, counts toward its step total. Because it's squared, extra steps tank the score fast.

So "mostly cleared" can mean wildly different scores depending on efficiency. The source gives a paired example:

Agent A · cleared and efficient
97.7%

Cleared all 7 levels in 785 steps total, nearly matching the human baseline of 776. Both cleared and efficient — near-perfect score.

Agent B · cleared but wasteful
<14%

Used 1,591 steps just to clear the first 6 levels — 2.7x the human baseline — and still hadn't cleared level 7. The squared penalty plus the incomplete-run cap pushed the score below 14.

So this 98.98 is a composite score of "how much got completed" and "how efficiently" — not a solve rate. It forces the AI not just to play the game, but to play it as economically as a skilled human.