OpenAI's highest-compute o1 variant that thinks harder for more consistent answers on difficult reasoning, math, and analysis tasks.
Output Speed *
Intelligence Index *
Context Window *
Input price
Output price
Performance
Why o1-pro Trades Speed for Answer Reliability
Benchmarks
o1-pro Reasoning and Reliability Signals
Output Speed
Intelligence Index
MMLU *
GPQA *
HLE *
LiveCodeBench *
Technical Specifications
What the model supports
Quickstart
import requestsurl = "https://api.anyapi.ai/v1/chat/completions"
payload = {
"model": "o1-pro",
"messages": [
{
"role": "user",
"content": [
{
"type": "text",
"text": "Text prompt"
},
{
"image_url": {
"url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
},
"type": "image_url"
}
]
}
]
}
headers = {
"Authorization": "Bearer AnyAPI_API_KEY",
"Content-Type": "application/json"
}
response = requests.post(url, json=payload, headers=headers)
print(response.json())
const url = 'https://api.anyapi.ai/v1/chat/completions';
const options = {
method: 'POST',
headers: {
Authorization: 'Bearer AnyAPI_API_KEY',
'Content-Type': 'application/json'
},
body: '{
"model":"o1-pro",
"messages":[
{
"role":"user",
"content":[
{
"type":"text",
"text":"Text prompt"
},
{
"image_url":{
"url":"https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
},
"type":"image_url"
}
]
}
]
}'
};
try {
const response = await fetch(url, options);
const data = await response.json();
console.log(data);
} catch (error) {
console.error(error);
}
curl --request POST \
--url https://api.anyapi.ai/v1/chat/completions \
--header 'Authorization: Bearer AnyAPI_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"model": "o1-pro",
"messages": [
{
"role": "user",
"content": [
{
"type": "text",
"text": "Text prompt"
},
{
"image_url": {
"url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
},
"type": "image_url"
}
]
}
]
}'
Comparison
o1-pro vs o1: When Is the Extra Compute Worth It?
o1-pro and standard o1 share the same core reasoning family, a 200,000-token context window, 100,000 max output tokens, an October 2023 knowledge cutoff, text and image input, and text output. The difference is compute: o1-pro spends more inference effort to think harder and return more consistent answers, and it reports higher math accuracy such as roughly 86% versus 78% on AIME. The practical decision is whether your workload's hardest questions justify materially higher cost and slower responses, or whether standard o1 already resolves them reliably enough.
Choose o1-pro when correctness on the hardest cases is the priority, retries are expensive, and you can tolerate slow, non-streaming responses through the Responses API, for example high-stakes research, competition-grade math, or deep code and document analysis. Choose standard o1 when it already answers your problems reliably, when you need Chat Completions compatibility or streaming, or when cost and latency matter and the marginal accuracy gain of o1-pro does not justify its substantially higher price.
Limitations & Trade-offs
Best-Fit Workloads
Where this model earns its place
Hard mathematical and scientific reasoning
o1-pro's central strength is difficult multi-step reasoning. Its higher AIME performance versus standard o1 and its focus on repeated-attempt reliability make it well suited to competition-grade math, quantitative modeling, and scientific problem solving where a single dependable answer is the goal. The trade-off is slow responses, so run these tasks asynchronously rather than interactively.
Deep code analysis and debugging
For subtle, cross-file debugging and complex logic that smaller models fragment, o1-pro's added compute helps it maintain coherence across long reasoning chains. Reviewers note it lets developers stop breaking problems into small pieces. Combined with the 200,000-token context window, it can analyze large code segments in one pass, though latency and cost make it better for hard cases than routine autocomplete-style coding.
Long-document research and analysis
The 200,000-token context window lets o1-pro reason over lengthy legal, financial, or technical documents in a single pass, supporting research assistants that must synthesize large source material carefully. Because responses are slow and outputs are billed with reasoning tokens included, this fits asynchronous, high-value analysis rather than high-throughput document pipelines.
Structured extraction from complex inputs
With JSON-schema structured outputs, o1-pro can produce reliably formatted results from difficult inputs where reasoning quality drives extraction accuracy, such as deducing relationships or resolving ambiguous requirements. Use structured outputs to constrain responses, and keep reasoning effort and output length bounded to control cost on these jobs.