OpenAI says GPT-6 Astra's low setting beats GPT-5.6 Sol's high.
It suggests switching to Astra to save tokens.
Token use on some tasks drops up to 3–4×.
What happened
OpenAI's product lead Tibo Sotiou posted on X on September 6 that GPT-6 Astra's 'low' setting outperforms GPT-5.6 Sol's 'high' setting. He suggested users satisfied with Sol's high setting switch to Astra's low or medium to save tokens.
Why it matters
Astra launched on September 3 and some users said it consumes a lot of tokens. Sotiou also announced that for flat-rate plan users, token usage on 'long-tail' tasks (likely some high-token tasks) is reduced by up to 3–4×.
What to watch
The claim hinges on actual user experience of Astra's low and medium settings versus Sol's high, especially for long-tail tasks. Watch whether token consumption indeed drops 3–4× as stated.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
OpenAI's new model, GPT-6 Astra, launched on September 3, and already the company is addressing user feedback about high token consumption. The product lead's comparison between Astra's low and Sol's high settings is a direct push to migrate users, implying that the new model offers better performance per token. This is a efficiency pitch to reassure users concerned about costs.
Sotiou's announcement of up to 3–4× less token usage on long-tail tasks for flat-rate plan users reinforces the cost-saving narrative. However, the claim's credibility rests on real-world performance—whether Astra's low truly matches or beats Sol's high across diverse tasks. If it holds, it could encourage broader adoption; if not, it might raise skepticism about OpenAI's messaging.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
JM-Applied, a Taiwanese semiconductor gas equipment maker and supply chain member for Micron and TSMC, said on…
Nvidia disclosed roughly $99 billion of public and private equity investments as of July 26, plus about $25 bi…
In January, Ukraine's defense ministry said it would share millions of data points from tens of thousands of d…
OpenAI Group PBC acknowledged it did not publicly disclose an episode where its AI agents wrote to outside web…
Eaton is expanding beyond traditional power management into modular power deployment, next-generation DC conve…
OpenAI published two blog posts on September 6: a research acceleration report and an essay by Chief Scientist…
