ResearchPublished All news from that day
OpenAI publishes LifeVal, a personal-assistant test drawn from time-use surveys
OpenAI's own benchmark covers tasks such as finding childcare and accessing government services. Its current models score 90.4% to 100% on the five example tasks it shows.
Check our sources · 5 facts from 1 sourceOpenAI says it built LifeVal to measure model performance on everyday tasks such as finding childcare, accessing government services, creating a budget and planning workouts. It published the evaluation on 2 November 2026.
To pick the tasks, OpenAI started from the American Time Use Survey and the Multinational Time Use Study, then added a new survey of 3,995 respondents on where people most want help from a personal assistant. It lists caregiving, health, education, household administration, personal finances, shopping and government services among the areas.
The post shows five example tasks with scores for OpenAI's current models, from 90.4% on a diabetes meal-prep plan to 100% on finding a family home. It says the latest models score very highly on many LifeVal tasks. The scores come from OpenAI's own evaluation of its own models.
Underlined sentences link to the facts we checked. Dotted words open a short explainer.
Fact 2 of 5 · from OpenAI, 2 Nov 2026
the American Time Use Survey
Your reaction
We count reactions per story and day, never who reacted. The counts help us choose what goes in the monthly issue. If you are signed in, your own page shows yours too.
We count reactions per story and day, never who reacted. They help choose what goes in the monthly issue.
Check our sources
Every sentence above is checked against this source.
1 LifeVal
Open the source archived copy-
OpenAI's LifeVal page is dated 2 November 2026 and describes an evaluation of tasks covering areas where users wanted the most help from AI and personal agents, such as finding childcare, accessing government services, creating a budget and designing workout plans.
citewe built an evaluation of tasks covering areas where users wanted the most help from AI and personal agents
-
OpenAI says it started from the American Time Use Survey and the Multinational Time Use Study and added a new demographically representative survey of 3,995 respondents.
citethe American Time Use Survey
-
OpenAI says the resulting tasks span caregiving, health, education, household administration, personal finances, shopping and accessing government services.
citeincluding caregiving, health, education, household administration, personal finances, shopping, and accessing government services
-
The LifeVal page shows five example tasks with scores of 93.1%, 98.0%, 100%, 96.3% and 90.4% for its current models, the lowest being the diabetes meal-prep task and the highest the home-finding task.
citeMeal prep for a parent adjusting to a new diabetes diagnosis 90.4%
-
OpenAI says its latest models score very highly on many LifeVal tasks. the company says
citeOur latest models score very highly on many LifeVal tasks.
Every sentence above is checked against this source. Each fact has its own address you can cite.
Source 1 · OpenAI · 2 Nov 2026
LifeVal
OpenAI's LifeVal page is dated 2 November 2026 and describes an evaluation of tasks covering areas where users wanted the most help from AI and personal agents, such as finding childcare, accessing government services, creating a budget and designing workout plans.
we built an evaluation of tasks covering areas where users wanted the most help from AI and personal agents
OpenAI says it started from the American Time Use Survey and the Multinational Time Use Study and added a new demographically representative survey of 3,995 respondents.
the American Time Use Survey
OpenAI says the resulting tasks span caregiving, health, education, household administration, personal finances, shopping and accessing government services.
including caregiving, health, education, household administration, personal finances, shopping, and accessing government services
Show all 5 factsShow fewer facts
The LifeVal page shows five example tasks with scores of 93.1%, 98.0%, 100%, 96.3% and 90.4% for its current models, the lowest being the diabetes meal-prep task and the highest the home-finding task.
Meal prep for a parent adjusting to a new diabetes diagnosis 90.4%
OpenAI says its latest models score very highly on many LifeVal tasks. the company says
Our latest models score very highly on many LifeVal tasks.
Topics
The morning email
On the mornings we publish: the three top stories and up to four short ones. Free.
We email you a link to confirm. An issue may include one sponsor, always labelled Sponsored · Advertisement. Our emails count opens and clicks, not who made them. Unsubscribe in one click. What we keep