# This is fun. I finally got to follow up on a RemindMe comment. Back on March 25 of this year, six months ago, no model was getting even 1% on ARC-AGI-3. A commenter asked if we could see 75% at $2 cost. Well, the cost is still high ($26.1k), but GPT-6-Astra-Max was able to get 62.7% (no harness!) — retained record

> Retained reddit record from reddit:singularity.

Canonical: https://highsignal.app/data/records/8eafd65b29963414

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:singularity)  published 2026-09-25 18:19 ingested 2026-09-26 02:37

[original source ↗](https://www.reddit.com/r/singularity/comments/1wq3qb4/this_is_fun_i_finally_got_to_follow_up_on_a/)

## retained text



```
And the commenter did technically say "a year from now", so to me that makes the 62.7% less than six months later (Astra was released earlier this month, on September 3), even more impressive.

And of course with a lightweight memory adapter*, the benchmark is simply saturated. 

Those saturation scores are still at high cost, at least 17k, but if the trend of cheaper intelligence continues (https://epoch.ai/publications/the-plunging-price-of-thought), we should have models beating ARC-AGI-3 cheaper soon enough.

"The provider adapter preserves opaque reasoning state between requests and uses compaction for longer conversations, so the model can reuse prior work rather than reconstructing it."
```



document fetched 2026-09-26 00:30 · canonical https://reddit.com/r/singularity/comments/1wq3qb4/this_is_fun_i_finally_got_to_follow_up_on_a
