Azure Direct Models

Direct from Azure models are a select portfolio curated for their market-differentiated capabilities:

Secure and managed by Microsoft: Purchase and manage models directly through Azure with a single license, consistent support, and no third-party dependencies, backed by Azure's enterprise-grade infrastructure.
Streamlined operations: Benefit from unified billing, governance, and seamless PTU portability across models hosted on Azure - all as part of one Azure AI Foundry platform.
Future-ready flexibility: Access the latest models as they become available, and easily test, deploy, or switch between them within Azure AI Foundry; reducing integration effort.
Cost control and optimization: Scale on demand with pay-as-you-go flexibility or reserve PTUs for predictable performance and savings.

Learn more about Direct from Azure models .

Key capabilities

About this model

Introducing our new multimodal AI model, which now supports both text and audio modalities.

Key model capabilities

Enhanced customer service: By integrating audio inputs, gpt-4o-realtime-preview enables more dynamic and comprehensive customer support interactions.
Content innovation: Use gpt-4o-realtime-preview's generative capabilities to create engaging and diverse audio content, catering to a broad range of consumer preferences.
Real-time translation: Leverage gpt-4o-realtime-preview's capability to provide accurate and immediate translations, facilitating seamless communication across different languages

Use cases

See Responsible AI for additional considerations for responsible use.

Key use cases

The introduction of gpt-4o-realtime-preview opens numerous possibilities for businesses in various sectors: Enhanced customer service, content innovation, and real-time translation capabilities facilitate seamless communication across different languages.

Out of scope use cases

Currently, the gpt-4o-realtime-preview model focuses on text and audio and does not support existing gpt-4o features such as image modality and structured outputs. For many tasks, the generally available gpt-4o models may still be more suitable. IMPORTANT: At this time, gpt-4o-realtime-preview usage limits are suitable for test and development. To prevent abuse and preserve service integrity, rate limits will be adjusted as needed. IMPORTANT: The system stores your prompts and completions as described in the "Data Use and Access for Abuse Monitoring" section of the service-specific Product Terms for Azure OpenAI Service, except that the Limited Exception does not apply. Abuse monitoring will be turned on for use of the GPT-4o-realtime-preview API even for customers who otherwise are approved for modified abuse monitoring.

Pricing

Pricing is based on a number of factors, including deployment type and tokens used. See pricing details here.

Technical specs

The provider has not supplied this information.

Training cut-off date

The provider has not supplied this information.

Training time

The provider has not supplied this information.

Input formats

Currently, the gpt-4o-realtime-preview model focuses on text and audio and does not support existing gpt-4o features such as image modality and structured outputs.

Output formats

The provider has not supplied this information.

Supported languages

The provider has not supplied this information.

Sample JSON response

The provider has not supplied this information.

Model architecture

The provider has not supplied this information.

Long context

The provider has not supplied this information.

Optimizing model performance

The provider has not supplied this information.

Additional assets

The following documents are applicable:

Training disclosure

Training, testing and validation

GPT-4o-realtime-preview has safety built-in by design across modalities, through techniques such as filtering training data and refining the model's behavior through post-training.

Distribution

Distribution channels

The provider has not supplied this information.

More information

We've evaluated GPT-4o-realtime-preview according to our Preparedness Framework and in line with our voluntary commitments . Our evaluations of cybersecurity, CBRN, persuasion, and model autonomy show that GPT-4o-realtime-preview does not score above Medium risk in any of these categories. This assessment involved running a suite of automated and human evaluations throughout the model training process. We tested both pre-safety-mitigation and post-safety-mitigation versions of the model, using custom fine-tuning and prompts, to better elicit model capabilities. GPT-4o-realtime-preview has also undergone extensive external red teaming with 70+ external experts in domains such as social psychology, bias and fairness, and misinformation to identify risks that are introduced or amplified by the newly added modalities. We used these learnings to build out our safety interventions in order to improve the safety of interacting with GPT-4o-realtime-preview. We will continue to mitigate new risks as they're discovered. Model Versions:

2024-12-17: Updating the gpt-4o-realtime-preview model with improvements in voice quality and input reliability. As this is a preview version, it is designed for testing and feedback purposes and is not yet optimized for production traffic.
2024-10-01: Introducing our new multimodal AI model, which now supports both text and audio modalities. As this is a preview version, it is designed for testing and feedback purposes and is not yet optimized for production traffic.

Responsible AI considerations

Safety techniques

GPT-4o-realtime-preview has safety built-in by design across modalities, through techniques such as filtering training data and refining the model's behavior through post-training. We have also created new safety systems to provide guardrails on voice outputs.

Safety evaluations

We've evaluated GPT-4o-realtime-preview according to our Preparedness Framework and in line with our voluntary commitments . Our evaluations of cybersecurity, CBRN, persuasion, and model autonomy show that GPT-4o-realtime-preview does not score above Medium risk in any of these categories. This assessment involved running a suite of automated and human evaluations throughout the model training process. We tested both pre-safety-mitigation and post-safety-mitigation versions of the model, using custom fine-tuning and prompts, to better elicit model capabilities. GPT-4o-realtime-preview has also undergone extensive external red teaming with 70+ external experts in domains such as social psychology, bias and fairness, and misinformation to identify risks that are introduced or amplified by the newly added modalities. We used these learnings to build out our safety interventions in order to improve the safety of interacting with GPT-4o-realtime-preview. We will continue to mitigate new risks as they're discovered.

Known limitations

IMPORTANT: The system stores your prompts and completions as described in the "Data Use and Access for Abuse Monitoring" section of the service-specific Product Terms for Azure OpenAI Service, except that the Limited Exception does not apply. Abuse monitoring will be turned on for use of the GPT-4o-realtime-preview API even for customers who otherwise are approved for modified abuse monitoring. Currently, the gpt-4o-realtime-preview model focuses on text and audio and does not support existing gpt-4o features such as image modality and structured outputs. For many tasks, the generally available gpt-4o models may still be more suitable. IMPORTANT: At this time, gpt-4o-realtime-preview usage limits are suitable for test and development. To prevent abuse and preserve service integrity, rate limits will be adjusted as needed.

Acceptable use

Acceptable use policy

The provider has not supplied this information.

Quality and performance evaluations

Source: OpenAI The provider has not supplied this information.

Benchmarking methodology

Source: OpenAI We've evaluated GPT-4o-realtime-preview according to our Preparedness Framework and in line with our voluntary commitments . Our evaluations of cybersecurity, CBRN, persuasion, and model autonomy show that GPT-4o-realtime-preview does not score above Medium risk in any of these categories. This assessment involved running a suite of automated and human evaluations throughout the model training process. We tested both pre-safety-mitigation and post-safety-mitigation versions of the model, using custom fine-tuning and prompts, to better elicit model capabilities. GPT-4o-realtime-preview has also undergone extensive external red teaming with 70+ external experts in domains such as social psychology, bias and fairness, and misinformation to identify risks that are introduced or amplified by the newly added modalities. We used these learnings to build out our safety interventions in order to improve the safety of interacting with GPT-4o-realtime-preview. We will continue to mitigate new risks as they're discovered.

Public data summary

Source: OpenAI The provider has not supplied this information.