OpenAI’s long-teased “Strawberry” project became public on September 12, 2024, under the official name OpenAI o1-preview, the first release in the o1 reasoning-model series. OpenAI positioned it as a model that spends more time reasoning before answering, especially on difficult math, coding, and science problems. The announcement was a 2024 launch, not a new release in 2026; current availability should be checked in OpenAI’s product interfaces and release notes.
What is OpenAI’s “Strawberry” model called?
“Strawberry” was the pre-release name associated with the project. The public model announced by OpenAI on September 12, 2024, was o1-preview, first in the o1 series. OpenAI also introduced o1-mini, a smaller, faster option aimed at reasoning tasks that do not require as much broad world knowledge.
As an Amazon Associate I earn from qualifying purchases.
OpenAI described o1-preview as an early preview that would be updated. The name “Strawberry” is not the public product name used in the launch announcement. For later model changes and retirements, consult OpenAI’s Model Release Notes.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →What was o1 designed to do?
OpenAI said the o1 models were trained to spend more time thinking before responding, with the goal of handling harder problems in science, coding, and mathematics than earlier models. The practical distinction is task fit: a reasoning-oriented model may be useful when a problem benefits from working through several steps, but that does not make it automatically better for every request.
OpenAI’s September 2024 launch post reported that its reasoning model scored 83% on a qualifying International Mathematics Olympiad exam, compared with 13% for GPT-4o, and reached the 89th percentile in Codeforces contests. These are results reported by OpenAI for its stated evaluations, not independent replications or a guarantee of performance on an individual user’s work.
In its research explainer, OpenAI also said human evaluators preferred o1-preview over GPT-4o in reasoning-heavy categories such as data analysis, coding, and math, but not for some natural-language tasks. That qualification is important: benchmark strength and preference depended on the category being evaluated.
How did o1-preview differ from GPT-4o at launch?
| Area | o1-preview and o1-mini at September 2024 launch | GPT-4o at that time |
|---|---|---|
| Intended use | o1-preview targeted harder reasoning tasks in math, coding, and science. o1-mini was positioned as a faster, less expensive reasoning option, particularly suited to coding tasks that did not need broad world knowledge. | OpenAI described GPT-4o as more capable for many common cases in the near term. |
| Reported evaluations | OpenAI reported 83% on its qualifying IMO exam for the reasoning model and the 89th percentile in Codeforces contests. | OpenAI reported 13% on the same qualifying IMO exam for GPT-4o. |
| Natural-language preference | Human evaluators preferred o1-preview in some reasoning-heavy categories, but not some natural-language tasks, according to OpenAI’s research explainer. | Preference depended on task category; the launch post cautioned that GPT-4o could be more capable for many common uses. |
| Tools and inputs | At launch, o1-preview did not support web browsing or file and image uploads in ChatGPT. | The launch announcement positioned GPT-4o as the more capable choice for many common cases, though it did not establish a complete feature-by-feature comparison here. |
| Cost and speed positioning | OpenAI said o1-mini was 80% cheaper than o1-preview at launch and described it as faster. This was an announcement-era comparison, not a current price quote. | Not stated as a comparable price or speed figure in the launch announcement. |
OpenAI summed up the trade-off in its launch announcement: “For many common cases GPT‑4o will be more capable in the near term.” That was the company’s guidance for the initial preview period, not a claim about every later model or today’s product lineup.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Could you use o1 in ChatGPT or through the API?
Yes, according to the launch-era announcement, but the access terms were specific to September 2024 and should not be read as current entitlements. OpenAI said ChatGPT Plus and Team users would get access starting on launch day, with Enterprise and Edu access planned for the following week. It also opened API prototyping to developers who met its usage-tier requirements, with a stated limit of 20 requests per minute at that time.
Those historical rollout details do not establish whether o1-preview, a later o1 model, or any particular feature is available to a given account now. OpenAI’s ChatGPT release notes describe product changes over time, and model and plan availability can change. Check the model picker or current official API documentation for live access and limits.
What were the preview’s limitations and safety claims?
At launch, o1-preview lacked web browsing and file or image uploads in ChatGPT. OpenAI also said GPT-4o could be more capable for many common tasks, and its evaluator results showed that o1-preview was not preferred for some natural-language work. The preview was therefore a specialized addition, not a drop-in replacement with every feature of a general-purpose assistant.
OpenAI said its Safety and Security Committee and Board reviewed the o1 safety assessment and launch readiness. Its September 2024 research explainer reported safe-completion scores of 0.934 for o1-preview and 0.714 for GPT-4o on challenging jailbreak and edge-case prompts. These are results from an OpenAI-run evaluation on specified prompts; they are not an independent safety certification or a general probability that any response will be safe.
Free tools Windows power users keep installed
One-click scans. No signup required.
What the “Strawberry” launch means now
The useful takeaway is historical and practical: “Strawberry” became OpenAI o1-preview, an early 2024 release designed to improve reasoning on selected difficult tasks. OpenAI’s launch figures support its case for those evaluations, while its own caveats make clear that the model was not superior for every task and did not initially have the feature breadth of GPT-4o. Because product names, access, and limits have since changed, use OpenAI’s current product surfaces rather than the 2024 rollout terms to determine what is available today.
Quick Recap
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.
