As of: 2026-11-13 13:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/m/gpt-realtime-2/

# GPT-Realtime-2

OpenAI, GPT Realtime family. Status: current. Type: speech and audio. Released: 2026-05-07. Training cutoff: 2024-09.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| OpenAI API | `gpt-realtime-2` | wss://api.openai.com/v1/realtime | https://developers.openai.com/api/docs/models/gpt-realtime-2 |

## Pricing

- Text input: $4 per 1M tokens
- Text output: $24 per 1M tokens
- Audio input: $32 per 1M tokens
- Audio output: $64 per 1M tokens
- Cached input: $0.40 per 1M tokens
- Image input: $5 per 1M tokens
- Unit as published: per 1M tokens (USD); cached audio/text input $0.40, cached image $0.50
- Source: https://developers.openai.com/api/docs/pricing
- Last checked: 2026-09-29

## Specs

- Input: text, audio, image
- Output: text, audio
- Context window: 128,000 tokens
- Max output: 32,000 tokens
- Open weights: no
- Licence: proprietary

## Where it stands

- Newer speech and audio models in the registry since its release: 57. The 5 newest:
  - NVIDIA Nemotron 3.5 Transcribe (30B-A4B) (2026-10-28) https://postcutoff.com/m/nemotron-3-5-transcribe/
  - Hy-SimulMT-30B-A3B (2026-10-22) https://postcutoff.com/m/hy-simulmt-30b-a3b/
  - Eleven Realtime v1 Max / Mini (2026-10-20) https://postcutoff.com/m/eleven-realtime-v1-max/
  - Voxtral Mini Transcribe 3 (+ Mini Transcribe Realtime 3) (2026-10-20) https://postcutoff.com/m/voxtral-mini-transcribe-3/
  - Voxtral Small Transcribe 3 (2026-10-20) https://postcutoff.com/m/voxtral-small-transcribe-3/
- Before it from OpenAI: GPT-Audio-1.5 (and gpt-audio / gpt-audio-mini) (2026-02-23) https://postcutoff.com/m/gpt-audio-1-5/
- After it from OpenAI: GPT-Realtime-2.1 (2026-07-06) https://postcutoff.com/m/gpt-realtime-2-1/
- Closed weights; access is hosted only.

## What stands out

- Reasoning speech-to-speech model: First OpenAI realtime voice model with configurable reasoning effort (press: 'GPT-5-class' reasoning); higher effort adds latency and tokens. Source: https://developers.openai.com/api/docs/models/gpt-realtime-2
- 128K-token realtime context: Context grew from 32K (gpt-realtime-1.5) to 128K tokens, with 32K max output, for long voice-agent sessions. Source: https://www.ghacks.net/2026/05/11/openai-releases-three-new-realtime-voice-models-for-the-api-with-gpt-5-class-reasoning/

## Notes

Launched 2026-05-07 with gpt-realtime-translate and gpt-realtime-whisper (changelog). Superseded two months later by gpt-realtime-2.1 (2026-07-06) at identical prices, but still listed and not deprecated. Realtime endpoint only; function calling and prompt caching. Official launch post (openai.com) returned 403 to our fetcher, so benchmark claims were not read directly; secondary sources quote OpenAI: +15.2% Big Bench Audio vs gpt-realtime-1.5 (high effort), +13.8% Audio MultiChallenge instruction following (xhigh); one blog reports 96.6% absolute Big Bench Audio at xhigh (unconfirmed).

OpenAI's first reasoning realtime voice model. For new builds prefer [gpt-realtime-2.1](gpt-realtime-2-1.md).

Sources:
- https://developers.openai.com/api/docs/models/gpt-realtime-2
- https://developers.openai.com/api/docs/pricing
- https://developers.openai.com/api/docs/changelog
- https://openai.com/index/advancing-voice-intelligence-with-new-models-in-the-api/

Model page: https://postcutoff.com/m/gpt-realtime-2/