AI News
Asana reports 76x cost cut and 5x speed gain in browser agent tests
OpenAI News reports Asana used GPT-6 Astra in Codex to make its browser agent 76x cheaper and 5x faster in tests. Read the practical implications.
What this piece is grounded in
01
Asana made its browser agent 76x cheaper and 5x faster in tests
02
Using GPT-6 Astra in Codex, Asana made its browser agent 76x cheaper
03
Asana cuts model costs 76x in browser tests with GPT-6.1 Sol
04
to offer customers more capable models
01 / FIELD NOTE
What changed
According to an OpenAI News RSS summary published on 2026-10-09, Asana used GPT-6 Astra in Codex to make its browser agent 76x cheaper and 5x faster in tests. The summary states the goal was to offer customers more capable models. The original article title is 'Asana cuts model costs 76x in browser tests with GPT-6.1 Sol'. The publisher-provided summary is the only content available in this feed entry.
02 / FIELD NOTE
Why it matters for practical AI work
Editorial interpretation: The reported 76x cost reduction and 5x speed improvement in browser tests suggest that model choice can dramatically affect operational economics for agentic workflows. If these test results hold in production, teams building browser-based automation could see lower per-task costs and faster completion times. The summary explicitly ties the gains to using GPT-6 Astra in Codex, indicating the integration approach matters. However, these are test results, not production benchmarks.
03 / FIELD NOTE
What a reader can verify
A reader can verify the publisher summary by visiting the canonical article at openai.com/index/asana-browser-agent. The summary states the publication date as 2026-10-09 and identifies OpenAI News as the publisher. The exact figures reported are 76x cheaper and 5x faster. The summary also names GPT-6 Astra in Codex as the model and tool combination used. No other technical details are present in the feed entry.
04 / FIELD NOTE
What remains unclear
The supplied feed entry is a concise summary; consult the canonical article for full context. The summary does not describe the test methodology, the number of test cases, the baseline model used for comparison, or the specific browser tasks involved. It also does not state whether the 76x and 5x figures are measured against a previous Asana model or an external baseline. The category is unspecified by the publisher in this feed record.
Questions readers ask
What specific numbers did Asana report in the browser tests?
According to the OpenAI News RSS summary, Asana made its browser agent 76x cheaper and 5x faster in tests. The summary attributes these gains to using GPT-6 Astra in Codex. The original article title references GPT-6.1 Sol. These figures come from the publisher-provided summary, not from a detailed technical report.
Which model and tool did Asana use to achieve these results?
The OpenAI News summary states Asana used GPT-6 Astra in Codex to achieve the reported improvements. The article title mentions GPT-6.1 Sol, which may be a related or different model designation. The summary does not explain the relationship between GPT-6 Astra and GPT-6.1 Sol. Consult the canonical article for clarification.