Record · corpus.blog/posts/01a07d5d-9664-71d6-a7c2-b0fbcfe32e19

How enabling two settings tripled our scores on the ARC-AGI-3 benchmark

OpenAI · published 29 July 2026

A record is what was published and who pointed at it. The text of the post is not held here: read it at the source, or at its Wayback capture.

Read the post at openai.com Wayback capture

3blogs citing

First cited: 16 Jul 2026
Most recent: 3 Sept 2026
Rank this month: not ranked
Links from this post: 13
Words captured: 1,252

Cited by

Blog

In the post

Date

marktechpost.com

OpenAI Releases GPT-6 Astra: A 1.05M-Context Computer-Use Model Gated Behind a ‘Critical’ Cyber Threshold Read at the source ↗

3 Sept 2026

quesma.com

Kimi K3 is Open, Opus 5 is Good, DeepSeek V4 Flash is Cheap: LLMs on Baba Is You Read at the source ↗

31 Jul 2026

saito.ai

Your LLM Works. Your Product Probably Doesn't Read at the source ↗

30 Jul 2026

quesma.com

Baba Is Solved by Fable 5 and GPT-5.6 Sol, but at what cost? Read at the source ↗

16 Jul 2026

Links from this post

Link

Host

https://arcprize.org/tasks/cd82

arcprize.org

https://arcprize.org/arc-agi/3

arcprize.org

https://cdn.openai.com/pdf/04d1d1e4-bc75-476a-97cf-49055cd98d31/cdc_proof.pdf

cdn.openai.com

https://docs.arcprize.org/methodology

docs.arcprize.org

https://huggingface.co/datasets/magic-sword/arc_agi_3_public_demo_human_testing

huggingface.co

https://arcprize.org/tasks

arcprize.org

https://www.twitch.tv/gpt_plays_pokemon

twitch.tv

https://www.twitch.tv/epochaiplays

twitch.tv

Baba Is Solved by Fable 5 and GPT-5.6 Sol, but at what cost? Read at the source ↗

quesma.com

https://x.com/emollick/status/2075950897029374334

x.com

Analyzing GPT-5.5 & Opus 4.7 with ARC-AGI-3 Read at the source ↗

arcprize.org

https://developers.openai.com/blog/responses-api

developers.openai.com

https://developers.openai.com/api/docs/guides/compaction

developers.openai.com