← Back to Voice models
OpenAINotable
GPT-Realtime-1.5
Released Feb 23, 2026
OpenAI's non-reasoning Realtime API speech-to-speech model for voice agents and customer support, still listed and not deprecated.
- Price per minute
- —
- Pricing
- Per 1M tokens: text $4 in / $0.40 cached / $16 out; audio $32 in / $0.40 cached / $64 out; image $5 in / $0.50 cached.
- Inputs
- audio, text, image
- Tool calling
- Yes
- Languages
- —
- Voices
- 10
- Big Bench Audio
- 81.4%
- Access
- Generally available
- Licence
- Proprietary
Strengths & weaknesses
Strengths
- Lower latency than the reasoning models: 0.81s time to first audio on AA
- Function calling and prompt caching
- Text output $16 per 1M tokens, vs $24 on the 2.x models
Weaknesses
- 32K context and 4,096 max output tokens
- Big Bench Audio 81.4% (AA), well behind the 2.x models
Notable: frontier-lab clause. The model page still calls it 'the best voice model for audio in, audio out'. Not deprecated, but older than the 2.x line; consider dropping it if only the newest generation is kept. AA S2S Index 70.3.
Links
API model strings
- openai
gpt-realtime-1.5