GPT-6 Astra in Practice: Tests, Use Cases, and Feedback

Dots run on GPT-6 Astra, each with its own browser and cloud computer; while the user is away they do only read-only research, change-making actions follow approval defaults, and saved passwords stay hidden from the model.

The poster says Sol has launched with performance close to Astra on most tasks at one-fifth the price, and argues OpenAI is waging a price war.

Each dot gets its own cloud computer and browser, connects to over 4,000 apps, works inside ChatGPT, Slack and Teams, and sends updates or joins voice calls.

The poster calls DevDay all hype, saying the new model does not beat Opus 5.5 or even Astra, half the demos still use Astra, and the Dot bot is limited to the US.

Merge Gateway says GPT-6.1 Sol ties GPT-6 Astra on DeepSWE v1.1 and scores 6.4 points above GPT-6 Sol, at one-fifth the price.

An Arabic-speaking user describes Dots, an AI agent powered by GPT-6 Astra that runs autonomously 24/7 on a cloud computer with apps installed and can code.

A user complains that OpenAI is slow to bring Astra to Chat mode, calling the internally near-unlimited Astra a monster and saying they would have built GPT-7 by now.

Someone mocks GPT-6 Sol as so bad that OpenAI pulled a minor Astra version out of the drawer, relabeled it GPT-6.1 Sol and shipped it within a week, still short of Astra level.

GPT-6.1 Sol at medium matches GPT-6 Sol at Max with a score of 48, while GPT-6.1 Sol at Max scores 52, just one point below GPT-6 Astra at Max.

The $500/month Pro 500 centers on Ultrafast 8x output for GPT-6 Astra in Codex, and the author calculates that saving 23 minutes of waiting per day covers the cost, while listing quota consumption and fatigue as trade-offs.

Benchmark results show GPT-6.1 Sol (max) scoring 52, tied with Sonnet 5.5 (xhigh) and GPT-6 Astra (xhigh).

OpenAI announced Ultrafast at 300 tokens per second in Codex, about 8x faster, launching first with GPT-6 Astra, along with removal of the 5-hour limit and reopening of Pro 200.