Gpt-6 astra dropped hallucination from 92% to 51% the same week jensen called it agi. pick the…
gpt-6 astra dropped hallucination from 92% to 51% the same week jensen called it agi. pick the metric you want to win on.
one of them is the model lying to you less and the other is a press conference.
Context
Artificial Analysis's article, dated 9 September 2026, reports a drop in hallucination rate from 92% to 51% at max effort on AA-Omniscience, against GPT-5.6 Sol, with accuracy up 4 points. OpenAI's own system card says its factuality figures come from de-identified ChatGPT conversations flagged for factual errors and should not be read as production hallucination rates. The Next Web, dated 7 September 2026, reports Jensen Huang writing on X that AGI has arrived.
The 92% and 51% are Artificial Analysis's metric, not an OpenAI-reported rate, and that article is dated after the note, so it is not dated support for 7 September. The figures were not found in OpenAI's card. Huang's statement is reported secondhand and is not a technical result. Same week is consistent with 3 to 7 September. The model lying less versus a press conference is the author's framing.
Related work
Watch next
- The metric definition on Artificial Analysis and OpenAI's own figures.
Sources
- Benchmarking GPT-6 Astra (Artificial Analysis, 9 Sep 2026)artificialanalysis.ai
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 7 September 2026 at 12:04 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →