o1
Verifiedo1 delivers deep reasoning across text, images, and files.
About o1
o1 uses an internal chain-of-thought process to break down difficult problems before producing a final answer. This design differs from standard next-token predictors by allocating additional compute to planning and verification steps. The approach supports both text-only and vision-enabled queries.
Its primary strengths lie in mathematics, scientific reasoning, and multi-step coding challenges. Users commonly apply it to research analysis, algorithm design, and technical troubleshooting where accuracy depends on careful intermediate reasoning.
Capabilities
Benchmarks & performance
Independent evaluation scores and measured speed.
Source: Artificial Analysis
How o1 compares
o1 (striped bar) vs other multimodal on intelligence, speed and price.
Intelligence
Artificial Analysis Intelligence Index · Higher is better · o1 ranks #47 of 88
Speed
Output tokens per second · Higher is better · o1 ranks #35 of 76
Price
USD per 1M output tokens · Lower is better · o1 ranks #145 of 155
Sources: Artificial Analysis (intelligence, speed) · OpenRouter (price).
Best for
Complex Scientific Research
o1 excels at mathematical and scientific problem solving by applying advanced multi-step reasoning to derive solutions from intricate hypotheses and data sets.
Multimodal Content Analysis
The model performs multimodal text and image analysis, enabling detailed interpretation of combined visual and textual inputs for tasks like report generation or design review.
Large-Scale Code Development
o1 supports code generation with debugging and file-based data interpretation, making it effective for maintaining and optimizing extensive codebases over long contexts.
Strengths & limitations
Strengths
- +Strong performance on complex reasoning tasks
- +Effective chain-of-thought processing
- +Seamless vision-language integration
- +Handles extended technical contexts
Limitations
- –Slower response times from internal deliberation
- –Higher usage cost
- –Less optimized for quick or creative tasks
Cost calculator
Estimate what o1 would cost for your usage.
Based on o1's $15.00/1M input · $60.00/1M output. Estimate only — actual cost varies by provider and caching.
Quick start
OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. Only the model slug changes between models.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const completion = await client.chat.completions.create({
model: "openai/o1",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);Model slug: openai/o1
Editor's verdict
o1 is OpenAI's proprietary multimodal with a 200K-token context window.
On independent testing it scores 30.7 on the Artificial Analysis Intelligence Index, running at roughly 105 tokens per second with about 25.62s to first token.
At $60.00 per 1M output tokens, it is premium-priced for its class.
It is available through OpenAI's API and aggregators like OpenRouter.
Best suited to strong performance on complex reasoning tasks and effective chain-of-thought processing.
Frequently asked questions
o1 supports a context window of 200000 tokens for processing extended inputs.
User reviews
Real, verified reviews from the community shape this model's rating.
Loading reviews…
Other o models
Sibling versions in the o family from OpenAI.
Similar models
Other multimodal worth comparing.