Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
ARC Prize Profile picture
@arcprize
Jul 18, 2025 • 10 tweets • 4 min read • Read on X
Scrolly
Today, we're announcing a preview of ARC-AGI-3, the Interactive Reasoning Benchmark with the widest gap between easy for humans and hard for AI

We’re releasing:
* 3 games (environments)
* $10K agent contest
* AI agents API

Starting scores - Frontier AI: 0%, Humans: 100% Image
Every game environment is novel, unique, and only requires core-knowledge priors

No language, trivia, or specialized knowledge is needed to beat the games

* Play: three.arcprize.org
* Compete: arcprize.org/arc-agi/3/
* Build: three.arcprize.org/docs
Your ability to efficiently adapt to novelty defines your intelligence, not your performance on a single-skill

Harder puzzles don’t prove smarter AI, but rather its ability to learn new rules does

ARC Prize exists to operationalize that insight
We create benchmarks that highlight the gap between human generalization and machine pattern‑matching

ARC‑AGI‑1 (2019) challenged deep‑learning

ARC‑AGI‑2 (2025) challenges static reasoning models Image
Agents are now the frontier. They perceive, plan, act, remember, adapt. Static puzzles aren’t equipped to grade that loop

We need interactive benchmarks that test world‑model building and long‑horizon planning under sparse feedback Image
Enter ARC‑AGI‑3

6 brand‑new games. Easy for humans, out of reach for today’s best AI models

3 games are live today, 3 will go live in August Image
ARC-AGI API ships today

Plug in any LLM, RL, or hybrid agent, train locally, test against our servers Image
o3 (left) and Grok 4 (right) replays below

spoiler: neither complete a single level
ARC-AGI-3 Preview games need to be pressure tested. We’re hosting a 30-day agent competition in partnership with @huggingface

We’re calling on the community to build agents (and win money!)

arcprize.org/competitions/a…Image
We're excited to share this preview of ARC-AGI-3. This is just the beginning

Visit to learn morearcprize.org/arc-agi/3

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with ARC Prize

ARC Prize Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @arcprize

ARC Prize Profile picture
Dec 11, 2025
A year ago, we verified a preview of an unreleased version of @OpenAI o3 (High) that scored 88% on ARC-AGI-1 at est. $4.5k/task

Today, we’ve verified a new GPT-5.2 Pro (X-High) SOTA score of 90.5% at $11.64/task

This represents a ~390X efficiency improvement in one year Image
We also verified that GPT-5.2 Pro (High) is SOTA for ARC-AGI-2, scoring 54.2% for $15.72/task

(Due to API timeouts, we were unable to reliably verify GPT 5.2 Pro X-High on ARC-AGI-2)

All verified GPT-5.2 family scores:
arcprize.org/leaderboardImage
ARC-AGI is achieving its 2019 goal to push AI beyond memorization towards efficient on-the-fly adaptation

Reasoning systems now show genuine fluid intelligence on simple tasks
Read 7 tweets
ARC Prize Profile picture
Dec 5, 2025
Announcing the ARC Prize 2025 Top Score & Paper Award winners

The Grand Prize remains unclaimed

Our analysis on AGI progress marking 2025 the year of the refinement loop Image
ARC Prize 2025 Paper Award Winners

1st / "Less is More: Recursive Reasoning with Tiny Networks" (TRM) / A. Jolicoeur-Martineau / $50k

2nd / "Self-Improving Language Models for Evolutionary Program Synthesis: A Case Study on ARC-AGI" (SOAR) / J. Pourcel et al. / $20k

3rd / "ARC-AGI Without Pretraining" / I. Liao et al. / $5kImage
ARC Prize 2025 Top Score Winners

1st / NVARC / 25.03% / $25k
2nd / the ARChitects / 16.53% / $10k
3rd / MindsAI @ Tufa Labs / 12.64% / $5k
4th / Lonnie / 6.67% / $5k
5th / Guillermo Barbadillo @ Veridas / 6.53% / $5k Image
Read 12 tweets
ARC Prize Profile picture
Nov 18, 2025
Gemini 3 models from @Google @GoogleDeepMind have made a significant 2X SOTA jump on ARC-AGI-2 (Semi-Private Eval)

Gemini 3 Pro:
31.11%, $0.81/task

Gemini 3 Deep Think (Preview):
45.14%, $77.16/task Image
Also, notable - Gemini 3 results on ARC-AGI-1 (Semi-Private Eval)

Gemini 3 Pro:
75.00%, $0.49/task

Gemini 3 Deep Think (Preview):
87.50%, $44.26/task Image
Frontier AI reasoning systems are now closing the complexity scaling gap between ARC-AGI-1 and ARC-AGI-2

This is surprising, as these same systems also make obvious mistakes on easy tasks (for humans) from ARC-AGI-1. We're not sure why and invite help from the community to study this phenomenon

Full solution logs are linked in last tweet
Read 8 tweets
ARC Prize Profile picture
Aug 15, 2025
Analyzing the Hierarchical Reasoning Model by @makingAGI

We verified scores on hidden tasks, ran ablations, and found that performance comes from an unexpected source

ARC-AGI Semi Private Scores:
* ARC-AGI-1: 32%
* ARC-AGI-2: 2%

Our 4 findings: Image
Finding #1: The "hierarchical" architecture had minimal performance impact when compared to a similarly sized transformer

A drop-in transformer comes within a few points without any hyperparameter optimization.

See our full post: arcprize.org/blog/hrm-analy…Image
Finding #2: The under-documented “outer loop” refinement process drove substantial performance gains

Iterating over data by refining it has a strong impact, as the jump from 1 (no refinement) to 2 (1 refinement) shows. Image
Read 7 tweets
ARC Prize Profile picture
Jul 10, 2025
Grok 4 (Thinking) achieves new SOTA on ARC-AGI-2 with 15.9%

This nearly doubles the previous commercial SOTA and tops the current Kaggle competition SOTA Image
On ARC-AGI-1, Grok 4 (Thinking) achieves 66.7% inline with the Pareto frontier for AI reasoning systems we reported last month Image
Thank you to the @xai team for working with us to validate Grok 4's score and inviting us to the watch the live stream Image
Read 4 tweets
ARC Prize Profile picture
Apr 22, 2025
o3 and o4-mini on ARC-AGI's Semi Private Evaluation

* o3-medium scores 53% on ARC-AGI-1
* o4-mini shows state-of-the-art efficiency
* ARC-AGI-2 remains virtually unsolved (<3%)

Through analysis we highlight differences from o3-preview and other model behavior Image
As mentioned before, OpenAI has confirmed that the version of o3 that was released last week is not the same version that we tested in December ‘24.

For more on this see the tweet below or the blog post

During testing, both o3 and o4-mini frequently failed to return outputs when run at “high” reasoning.

The partial results we did receive are in the blog post.

However, these reasoning efforts were excluded from the leaderboard due to insufficient coverage.
Read 8 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

Send Email!

:(