What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The “as soon as this week” forecast was about September 2024, not a future launch: OpenAI announced o1-preview in ChatGPT and its API on September 12, 2024. Bloomberg reported that the release called o1 was connected to OpenAI’s internal “Strawberry” project, but OpenAI’s announcement named the public model series o1 and did not confirm that identity. “Human-like reasoning” was headline language, not a demonstrated equivalence to human reasoning.
What happened to the “as soon as this week” forecast?
On September 12, 2024, Bloomberg reported that a limited release of a model known internally as “Strawberry” could arrive as soon as that week. That same day, OpenAI announced o1-preview in ChatGPT and its API. Bloomberg connected o1 with the internal project name; the timing forecast was therefore followed by the public o1 announcement, but the name link was not confirmed in OpenAI’s announcement. Reuters’ earlier account and Bloomberg’s September 12 report provide the reporting context.
What did “Strawberry” mean?
Reuters reported on July 12, 2024, citing a person familiar with the matter and internal documents it reviewed, that OpenAI was working on a project code-named Strawberry. Reuters said the project had formerly been known as Q*, attributing that lineage to its reporting. The internal document described intended research use involving long-horizon tasks and autonomous web navigation.
Those reports did not establish that Strawberry was ready for public release or that an internal demo Reuters had heard about was Strawberry. Reuters also could not determine what was in the deep-research dataset cited in its report. An OpenAI spokesperson responded generally: “We want our AI models to see and understand the world more like we do.” The spokesperson added that continued research into AI capabilities was common in the industry and reflected a shared belief that systems would improve in reasoning over time. The statement did not directly confirm the reported Strawberry details. Reuters’ July 2024 report is the source for those details and qualifications.
#1 Best Overall
What did OpenAI say o1 could do?
OpenAI described o1 as the first model in a new series trained to spend more time thinking before answering. Its September 12 announcement said the models could refine their thinking, try strategies, and recognize mistakes. OpenAI identified science, coding, math, and multi-step workflows as potential use cases. These were the company’s descriptions, not independent findings about a model called Strawberry. OpenAI’s o1-preview announcement gives its launch-era explanation.
The announcement included several company-reported results, each tied to a particular evaluation rather than a general measure of reasoning:
- OpenAI said a reasoning model then in development scored 83% on an International Mathematics Olympiad qualifying exam, compared with 13% for GPT-4o. The 83% figure referred to the next model update under development, not a separately established result for Strawberry.
- OpenAI said its coding performance reached the 89th percentile in Codeforces competitions.
- On one of OpenAI’s hardest jailbreak tests, the company reported a score of 84 for o1-preview and 22 for GPT-4o on a 0–100 scale. That result was specific to the stated test; it was not a general safety guarantee.
These were OpenAI’s own reported evaluations. They do not establish human-level or human-like reasoning, and the figures should not be assigned to a separately verified Strawberry model.
What were o1-preview’s launch-era limitations and access terms?
OpenAI called o1-preview an early model. At launch, it said the model lacked ChatGPT features including web browsing and uploading files or images, and that GPT-4o might be more capable for many common uses in the near term. Those statements describe the September 2024 release, not current product availability.
Rank #3
OpenAI’s September 12 announcement said Plus and Team users could access o1 models in ChatGPT immediately, with Enterprise and Edu access planned for the following week. API prototyping was initially limited to qualifying tier-5 developers. The initial ChatGPT limits were 30 weekly messages for o1-preview and 50 for o1-mini; eligible developers had an API limit of 20 requests per minute. OpenAI updated its page on September 17, 2024, to list 50 weekly queries for o1-preview and 50 daily queries for o1-mini. These are historical launch-era limits, not present-day access guidance.
How did o1-mini differ from o1-preview?
OpenAI presented o1-mini as a smaller, faster model particularly effective at coding. In the launch announcement, the company said o1-mini was 80% cheaper than o1-preview. That was a September 2024 relative-price claim, not current pricing.
| Launch-era option | What OpenAI said | Access or limit stated at launch |
|---|---|---|
| o1-preview | Early reasoning model; suited to harder tasks, with the limitations described above. | Initially 30 weekly ChatGPT messages; updated September 17, 2024, to 50 weekly queries. Eligible API developers initially had a limit of 20 requests per minute. |
| o1-mini | Smaller, faster, and particularly effective at coding; OpenAI said it was 80% cheaper than o1-preview. | Initially 50 weekly ChatGPT messages; updated September 17, 2024, to 50 daily queries. |
All comparisons and limits in the table refer to OpenAI’s September 2024 launch announcement and its September 17 update; they are not current product terms. OpenAI’s announcement is the source.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Was Strawberry officially released under that name?
The public announcement was for OpenAI o1, not a product officially named Strawberry. Reuters established that Strawberry was a reported internal project, and Bloomberg reported a connection between the project and o1’s release. The available statements do not independently confirm that the two names referred to precisely the same model or disclose the internal project’s exact role. The careful conclusion is that OpenAI released o1-preview in September 2024 and contemporaneous reporting associated that release with Strawberry.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsQuick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
