diff --git a/docs/models/codex.mdx b/docs/models/codex.mdx index 49f3a5b2..6f39f717 100644 --- a/docs/models/codex.mdx +++ b/docs/models/codex.mdx @@ -21,7 +21,7 @@ OpenAI's [Codex CLI](https://github.com/openai/codex) is the same coding agent t ## Deploy a model in Foundry 1. Go to [Foundry](https://ai.azure.com) and create a new project. -1. From the [model catalog](https://ai.azure.com/catalog/) select a [reasoning model](/models/reasoning) such as `gpt-6-astra`, `gpt-5.6-sol`, `gpt-5.3-codex`, `gpt-5.2-codex`, `gpt-5.1-codex-max`, `gpt-5.1-codex`, `gpt-5.1-codex-mini`, [`gpt-5-codex`](https://ai.azure.com/catalog/models/gpt-5-codex), [`gpt-5`](https://ai.azure.com/catalog/models/gpt-5), [`gpt-5-mini`](https://ai.azure.com/catalog/models/gpt-5-mini), or [`gpt-5-nano`](https://ai.azure.com/catalog/models/gpt-5-nano). +1. From the [model catalog](https://ai.azure.com/catalog/) select a [reasoning model](/models/reasoning) such as `gpt-6-astra`, `gpt-6-sol`, `gpt-5.3-codex`, `gpt-5.2-codex`, `gpt-5.1-codex-max`, `gpt-5.1-codex`, `gpt-5.1-codex-mini`, [`gpt-5-codex`](https://ai.azure.com/catalog/models/gpt-5-codex), [`gpt-5`](https://ai.azure.com/catalog/models/gpt-5), [`gpt-5-mini`](https://ai.azure.com/catalog/models/gpt-5-mini), or [`gpt-5-nano`](https://ai.azure.com/catalog/models/gpt-5-nano). 1. To deploy the model from the model catalog, select **Use this model**. If you're using the Azure OpenAI **Deployments** pane, select **deploy model**. 1. Copy the endpoint **URL** and the **API Key**. @@ -52,7 +52,7 @@ From the terminal, run the following commands to install [Codex CLI](https://git ### Validated versions for GPT-6 Astra -Codex CLI versions `0.152.1` and `0.153.0` are validated with `gpt-6-astra` on Azure OpenAI, including [multi-agent orchestration](/models/responses-multi-agent) and [prompt caching](/models/prompt-caching). These versions are currently validated versions, not minimum supported versions. +Codex CLI versions `0.152.1` and `0.153.0` are validated with `gpt-6-astra` on Azure OpenAI, including [multi-agent orchestration](/models/responses-multi-agent) and [prompt caching](/models/prompt-caching). Codex CLI version `0.155.1` is validated with `gpt-6-sol` and `gpt-6-luna` on Azure OpenAI. These versions are currently validated versions, not minimum supported versions. ## Create and configure config.toml diff --git a/docs/models/model-retirement-schedule.mdx b/docs/models/model-retirement-schedule.mdx index 9e9dc4f0..ea2a3257 100644 --- a/docs/models/model-retirement-schedule.mdx +++ b/docs/models/model-retirement-schedule.mdx @@ -71,7 +71,7 @@ This section lists the retirement lifecycle for Foundry Models sold by Azure. | gpt-image-2.5-sunburst | 2026-09-09 | GA | 2027-09-09 | — | | gpt-realtime | 2025-08-28 | GA | 2027-03-02 | — | | gpt-realtime-1.5 | 2026-02-23 | GA | 2027-08-24 | — | -| gpt-realtime-2 | 2026-05-06 | Preview | 2026-08-31 | — | +| gpt-realtime-2 | 2026-05-06 | Preview | — | gpt-realtime-2.1 | | gpt-realtime-2.1 | 2026-07-07 | GA | 2027-06-25 | — | | gpt-realtime-2.1-mini | 2026-07-07 | GA | 2027-06-25 | — | | gpt-realtime-mini | 2025-10-06 | GA | 2027-04-06 | — | diff --git a/docs/models/models-sold-directly-by-azure-region-availability.mdx b/docs/models/models-sold-directly-by-azure-region-availability.mdx index 7ddc6eea..e14fc220 100644 --- a/docs/models/models-sold-directly-by-azure-region-availability.mdx +++ b/docs/models/models-sold-directly-by-azure-region-availability.mdx @@ -72,8 +72,8 @@ For **global deployments**, Azure OpenAI can process prompts and responses in an | gpt-5.6-sol | 2026-07-09 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | gpt-5.6-terra | 2026-07-09 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | gpt-6-astra | 2026-09-03 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | - | gpt-6-luna | 2026-09-22 | ✅ | ✅ | ✅ | ✅ | ✅ | - | ✅ | ✅ | ✅ | ✅ | - | gpt-6-sol | 2026-09-22 | ✅ | ✅ | ✅ | ✅ | ✅ | - | ✅ | ✅ | ✅ | ✅ | + | gpt-6-luna | 2026-09-22 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | + | gpt-6-sol | 2026-09-22 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | gpt-audio | 2025-08-28 | - | - | - | ✅ | - | ✅ | - | - | - | - | | gpt-audio-1.5 | 2026-02-23 | - | - | - | ✅ | - | ✅ | - | - | - | - | | gpt-audio-mini | 2025-10-06 | - | - | - | ✅ | - | ✅ | - | - | - | - | @@ -551,6 +551,8 @@ For **Data Zone** deployments, Microsoft processes prompts and responses anywher | gpt-5.6-sol | 2026-07-09 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | gpt-5.6-terra | 2026-07-09 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | gpt-6-astra | 2026-09-03 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | + | gpt-6-luna | 2026-09-22 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | + | gpt-6-sol | 2026-09-22 | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | | gpt-image-1.5 | 2025-12-16 | - | - | ✅ | - | - | - | ✅ | | model-router | 2025-05-19 | - | - | ✅ | - | - | - | ✅ | | model-router | 2025-08-07 | - | - | ✅ | - | - | - | ✅ | diff --git a/docs/models/models-sold-directly-by-azure.mdx b/docs/models/models-sold-directly-by-azure.mdx index a9a8ab88..9c301330 100644 --- a/docs/models/models-sold-directly-by-azure.mdx +++ b/docs/models/models-sold-directly-by-azure.mdx @@ -496,7 +496,7 @@ Details about maximum request tokens and training data are available in the foll |`gpt-realtime-whisper` (2026-05-06) | Audio model for real-time low-latency transcription. |Input: 32,000
Output: 4,096 | September 2024 | |`gpt-live-transcribe` (2026-07-29) | Audio model for real-time low-latency transcription. Current recommended model for realtime transcription scenarios. |Input: 32,000
Output: 4,096 | September 2024 | |`gpt-realtime-2` (2026-05-07) | Audio model for real-time audio processing. |Input: 32,000
Output: 4,096 | September 2024 | -|`gpt-realtime-2.1` (2026-07-07)
`gpt-realtime-2.1-mini` (2026-07-07)
**preview** | Audio models for real-time audio processing. Minor updates over `gpt-realtime-2` with improved silence and noise handling. |Input: 32,000
Output: 4,096 | September 2024 | +|`gpt-realtime-2.1` (2026-07-07)
`gpt-realtime-2.1-mini` (2026-07-07) | Audio models for real-time audio processing. `gpt-realtime-2.1` is an incremental update over `gpt-realtime-2` with improved silence and noise handling. `gpt-realtime-2.1-mini` is a smaller variant. |Input: 32,000
Output: 4,096 | September 2024 | diff --git a/docs/models/realtime-2.mdx b/docs/models/realtime-2.mdx index c793adec..88c9c70e 100644 --- a/docs/models/realtime-2.mdx +++ b/docs/models/realtime-2.mdx @@ -1,17 +1,20 @@ --- -title: "GPT Realtime 2.x (preview) overview" -description: "Learn about GPT Realtime 2, a speech-to-speech model with built-in reasoning for low-latency interactive voice experiences." +title: "GPT Realtime 2.x overview" +description: "Learn about the generally available GPT Realtime 2.1 models for low-latency, interactive voice experiences." --- -The GPT Realtime 2.x series models are speech-to-speech models with built-in reasoning. They accept audio input and produce audio output. They're designed for low-latency, interactive voice experiences where you need stronger instruction following and reasoning than earlier realtime models. +The generally available GPT Realtime 2.x series models are speech-to-speech models with built-in reasoning. They accept audio input and produce audio output. They're designed for low-latency, interactive voice experiences where you need stronger instruction following and reasoning than earlier realtime models. - +The current models are: -This feature is currently in public preview. This preview is provided without a service-level agreement, and we don't recommend it for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see [Supplemental Terms of Use for Microsoft Azure Previews](https://azure.microsoft.com/support/legal/preview-supplemental-terms/). +- `gpt-realtime-2.1` (version `2026-07-07`) is an incremental update to `gpt-realtime-2` with improved silence and noise handling. +- `gpt-realtime-2.1-mini` (version `2026-07-07`) is a new smaller variant in the 2.x series. - +Both models support Global Standard and Data Zone Standard deployments. For region-by-region availability, see [Region availability for Foundry Models sold by Azure](/models/models-sold-directly-by-azure-region-availability). -## What's new in GPT Realtime 2.x +Azure OpenAI bills these models by input, cached input, and output tokens. Image input is billed separately. For current rates, see the **Audio Models** section on the [Azure OpenAI pricing page](https://azure.microsoft.com/pricing/details/cognitive-services/openai-service/). + +## GPT Realtime 2.x features - **Reasoning support** with an adjustable `reasoning.effort` control. - **Response phases** that distinguish preambles ("commentary") from the final answer ("final_answer"). @@ -19,6 +22,22 @@ This feature is currently in public preview. This preview is provided without a ## Key concepts +### API and transports + +Use the GA Realtime API under `/openai/v1`. The WebSocket endpoint has the following format, where the `model` value is your deployment name: + +```text +wss://.openai.azure.com/openai/v1/realtime?model= +``` + +The GA endpoint doesn't use a date-based API version or an `api-version` query parameter. For complete examples, see [Use the GPT Realtime API via WebSockets](/models/realtime-audio-websockets) and [Use the GPT Realtime API via WebRTC](/models/realtime-audio-webrtc). + +### Audio and voices + +For audio sent as Realtime API events, use headerless, mono PCM16 audio sampled at 24 kHz and base64-encode the bytes. WebRTC negotiates audio transport between the client and service. + +The supported output voices are `alloy`, `ash`, `ballad`, `coral`, `cedar`, `echo`, `marin`, `sage`, `shimmer`, and `verse`. + ### Reasoning effort Control reasoning intensity with the `reasoning.effort` session parameter. Valid values are `minimal`, `low`, `medium`, and `high`. @@ -42,6 +61,8 @@ Instruction following is stricter than in earlier realtime models. If your syste Currently, GPT Realtime 2.x models don't support the `truncation` property in the `session.update` payload. To manage token usage and costs in an ongoing conversation, consider using the `conversation.item.truncate` or `conversation.item.delete` events. +Realtime models have separate audio-token and concurrent-session quotas. Review [Azure OpenAI quotas and limits](/models/quotas-limits) before production deployment. + ## Get started The connection and usage patterns for GPT Realtime 2.x are the same as for earlier versions. Deploy the new model and point your existing code to it. Choose the transport that fits your scenario: diff --git a/docs/models/realtime-audio-sip.mdx b/docs/models/realtime-audio-sip.mdx index 2c985d24..7b808583 100644 --- a/docs/models/realtime-audio-sip.mdx +++ b/docs/models/realtime-audio-sip.mdx @@ -19,6 +19,8 @@ The GPT real-time models are available for global deployments in [East US 2 and - `gpt-realtime-mini` (`2025-12-15`) - `gpt-realtime-1.5` (`2026-02-23`) - `gpt-realtime-2` (`2026-05-07`) +- `gpt-realtime-2.1` (`2026-07-07`) +- `gpt-realtime-2.1-mini` (`2026-07-07`) - `gpt-realtime-translate` (`2026-05-06`) - `gpt-realtime-whisper` (`2026-05-06`) - `gpt-live-transcribe` (`2026-07-29`) diff --git a/docs/models/realtime-audio-webrtc.mdx b/docs/models/realtime-audio-webrtc.mdx index 3e1b2aab..ebb04ca8 100644 --- a/docs/models/realtime-audio-webrtc.mdx +++ b/docs/models/realtime-audio-webrtc.mdx @@ -29,6 +29,8 @@ You can access the GPT real-time models for global deployments in the [East US 2 - `gpt-realtime-mini` (version 2025-12-15) - `gpt-realtime-1.5` (version 2026-02-23) - `gpt-realtime-2` (version 2026-05-07) +- `gpt-realtime-2.1` (version 2026-07-07) +- `gpt-realtime-2.1-mini` (version 2026-07-07) - `gpt-realtime-translate` (version 2026-05-06) - `gpt-realtime-whisper` (version 2026-05-06) - `gpt-live-transcribe` (version 2026-07-29) @@ -112,7 +114,7 @@ The session configuration includes: | `session.type` | Yes | Must be `realtime` | | `session.model` | Yes | Your model deployment name | | `session.instructions` | No | System prompt for the assistant | -| `session.audio.output.voice` | No | Voice for audio output: `alloy`, `ash`, `ballad`, `coral`, `echo`, `sage`, `shimmer`, or `verse` | +| `session.audio.output.voice` | No | Voice for audio output: `alloy`, `ash`, `ballad`, `coral`, `cedar`, `echo`, `marin`, `sage`, `shimmer`, or `verse` | ```python from flask import Flask, jsonify diff --git a/docs/models/realtime-audio-websockets.mdx b/docs/models/realtime-audio-websockets.mdx index 81e183c3..180632c9 100644 --- a/docs/models/realtime-audio-websockets.mdx +++ b/docs/models/realtime-audio-websockets.mdx @@ -2558,6 +2558,8 @@ The GPT real-time models are available for global deployments in the [East US 2 - `gpt-realtime-mini` (`2025-12-15`) - `gpt-realtime-1.5` (`2026-02-23`) - `gpt-realtime-2` (`2026-05-07`) +- `gpt-realtime-2.1` (`2026-07-07`) +- `gpt-realtime-2.1-mini` (`2026-07-07`) - `gpt-realtime-translate` (`2026-05-06`) - `gpt-realtime-whisper` (`2026-05-06`) - `gpt-live-transcribe` (`2026-07-29`) diff --git a/docs/models/realtime-audio.mdx b/docs/models/realtime-audio.mdx index b690e3aa..8c3bbc4e 100644 --- a/docs/models/realtime-audio.mdx +++ b/docs/models/realtime-audio.mdx @@ -34,6 +34,8 @@ The GPT real-time models are available for global deployments. - `gpt-realtime-mini` (version `2025-12-15`) - `gpt-realtime-1.5` (`2026-02-23`) - `gpt-realtime-2` (`2026-05-07`) +- `gpt-realtime-2.1` (`2026-07-07`) +- `gpt-realtime-2.1-mini` (`2026-07-07`) - `gpt-realtime-translate` (`2026-05-06`) - `gpt-realtime-whisper` (`2026-05-06`) - `gpt-live-transcribe` (`2026-07-29`) diff --git a/manifest.json b/manifest.json index 54d5d318..05dae21c 100644 --- a/manifest.json +++ b/manifest.json @@ -2741,7 +2741,7 @@ ] }, { - "name": "GPT Realtime 2.x (preview)", + "name": "GPT Realtime 2.x", "source_path": "articles/foundry/openai/concepts/realtime-2.md", "category": "in-repo-md", "section": "Models", @@ -2749,7 +2749,7 @@ "Models", "Model capabilities", "Realtime API", - "GPT Realtime 2.x (preview)" + "GPT Realtime 2.x" ], "output_path": "models/realtime-2", "includes": [ @@ -9416,7 +9416,7 @@ ] }, { - "name": "GPT Realtime 2.x (preview)", + "name": "GPT Realtime 2.x", "source_path": "articles/foundry/openai/concepts/realtime-2.md", "category": "in-repo-md", "section": "Models", @@ -9424,7 +9424,7 @@ "Models", "Model capabilities", "Realtime API", - "GPT Realtime 2.x (preview)" + "GPT Realtime 2.x" ], "output_path": "models/realtime-2", "includes": [