o3-deep-research
o3-deep-research is a specialized variant of o3 designed for deep research tasks, capable of extended multi-step investigation, source synthesis, and comprehensive analysis on complex questions that require thorough exploration.
import { streamText } from 'ai'
const result = streamText({ model: 'openai/o3-deep-research', prompt: 'Why is the sky blue?'})Playground
Try out o3-deep-research by OpenAI. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.
Providers
Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.
P50 throughput on live AI Gateway traffic, in tokens per second (TPS). Visit the docs for more info.
P50 time to first token (TTFT) on live AI Gateway traffic, in milliseconds. View the docs for more info.
Direct request success rate on AI Gateway and per-provider. Visit the docs for more info.
More models by OpenAI
| Model |
|---|
| 1M | 3.2s | 68tps | $5.00/M | $30.00/M | Read: $0.5/M Write: — | $10.00/K + input costs | — |   | 04/24/2026 | ||||
| 400K | 1.5s | 190tps | $0.75/M | $4.50/M | Read:$0.07/M Write:— | $10.00/K + input costs | — | | 03/17/2026 | ||||
| 400K | 0.5s | 117tps | $0.20/M | $1.25/M | Read:$0.02/M Write:— | $10.00/K + input costs | — | | 03/17/2026 | ||||
| 128K | 0.5s | 111tps | $1.25/M | $10.00/M | Read:$0.13/M Write:— | $10.00/K + input costs | — | | 11/12/2025 | ||||
| 400K | 3.4s | 446tps | $0.25/M | $2.00/M | Read:$0.03/M Write:— | $14/K + input costs | — | | 08/07/2025 | ||||
| 131K | 0.1s | 223tps | $0.35/M | $0.75/M | Read:$0.25/M Write:— | — | — |    +5 | 08/05/2025 |
About o3-deep-research
o3-deep-research was released on June 26, 2024 as a specialized variant of the o3 reasoning model, designed specifically for deep research tasks. While standard o3 applies chain-of-thought reasoning to produce answers, o3-deep-research extends this approach to conduct multi-step investigation, exploring questions from multiple angles, synthesizing information from diverse sources, and building comprehensive analyses.
The model spends significantly more time per request than standard models, reflecting the depth of investigation it performs. Each response represents the result of an extended research process rather than a single-pass generation. This makes it unsuitable for real-time applications but valuable for tasks where thoroughness and comprehensiveness matter.
If you need comprehensive answers to complex questions, o3-deep-research automates the kind of deep investigation that previously required hours of manual research and synthesis.
What To Consider When Choosing a Provider
- Configuration: o3-deep-research is designed for tasks that require thorough research rather than quick answers. It spends significantly more time per request, conducting multi-step investigation before producing results.
- Configuration: This is not a general chat model. It's built for deep analytical questions that benefit from extended exploration and synthesis.
- Zero Data Retention: AI Gateway does not currently support Zero Data Retention for this model. See the documentation for models that support ZDR.
- Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
When to Use o3-deep-research
Best For
- Literature review: Comprehensive analysis of research topics with source synthesis
- Due diligence research: Thorough investigation of companies, technologies, or markets
- Complex question answering: Questions that require exploring multiple angles and synthesizing diverse information
- Investigative analysis: Multi-step reasoning that builds on intermediate findings
- Report generation: Producing comprehensive reports that require deep exploration of a topic
Consider Alternatives When
- Quick answers: Standard o3 for reasoning tasks that don't require extended investigation
- General chat: GPT-5 or GPT-5.2 for conversational interactions
- Real-time responses: Any standard model when response time is constrained
- Simple lookups: Search-augmented models for straightforward information retrieval
Conclusion
o3-deep-research brings automated deep research capability to AI Gateway, conducting the kind of thorough, multi-step investigation that complex questions demand. For analytical workloads where comprehensiveness matters more than speed, it provides a level of depth that standard models cannot match.
Frequently Asked Questions
How does o3-deep-research differ from standard o3?
It conducts extended multi-step investigation rather than applying standard chain-of-thought reasoning. Responses take longer but are more comprehensive, synthesizing information from multiple angles.
How long do responses take?
Significantly longer than standard models, potentially minutes rather than seconds. The extended time reflects the depth of investigation performed.
What types of research questions work best?
Complex questions that benefit from exploring multiple perspectives, synthesizing diverse sources, and building comprehensive analyses. Simple factual queries don't need this model.
How does AI Gateway handle authentication for o3-deep-research?
AI Gateway accepts a single API key or OIDC token for all requests. You don't embed OpenAI credentials in your application; AI Gateway routes and authenticates on your behalf.
Can I use o3-deep-research for real-time chat?
No. Its extended processing time makes it unsuitable for real-time applications. Use it for background research tasks where thoroughness outweighs response time.
What are typical response times?
This page shows live performance metrics. Expect significantly longer response times than standard models due to the extended research process.