GPT-Realtime-2.1 pricing, context window and limits
GPT-Realtime-2.1 costs $4 per million input tokens and $24 per million output tokens at OpenAI list price. It accepts up to 128,000 tokens of context and returns up to 32,000 tokens.
- Model id
gpt-realtime-2.1 - Snapshot
What does GPT-Realtime-2.1 cost?
| Input | $4 |
|---|---|
| Output | $24 |
| Cached input (read) | $0.4 |
| Cache write | — |
US dollars per million tokens. A dash means no price is published.
How much can GPT-Realtime-2.1 read and write?
| Context window | 128,000 |
|---|---|
| Max input | 96,000 |
| Max output | 32,000 |
What can GPT-Realtime-2.1 do?
| Vision (image input) | Yes |
|---|---|
| Tool calling | Yes |
| Structured output | No |
| Reasoning | Yes |
| Input | text, audio, image |
| Output | text, audio |
What does a request to GPT-Realtime-2.1 cost?
| A typical request: 1,000 tokens in, 500 out | $0.016 |
|---|---|
| A long request: 10,000 tokens in, 1,000 out | $0.064 |
| 1,000 typical requests | $16 |
| 1 million tokens in and 1 million out | $28 |
Input read from the provider's prompt cache costs $0.4 per million tokens instead of $4, 90% less. A request that repeats a 1,000-token prompt from the cache costs $0.012 instead of $0.016.
Among 49 OpenAI models with a published price, GPT-Realtime-2.1 is the 32nd cheapest by input price. Its context window is the 34th largest of 48. List prices only; discounts, batch prices and prices above 200K tokens are not shown.
How do I call GPT-Realtime-2.1 through DVARA?
Configure the openai provider in the DVARA LLM Gateway and send the model id gpt-realtime-2.1. DVARA, the AI governance platform, applies your policies, audit and cost attribution to the call. See LLM cost attribution and budgets.
In DVARA, if you haven't set your own price for a model, its calls are costed at the price in the catalogue that ships with your DVARA version. A price you set in Flightdeck always wins. You can switch catalogue prices off. Then a model with no price you set is counted as costing nothing. Every cost record says which price it used.
Which other OpenAI models are there?
- gpt-image-2$5 in · $30 out · — context
- gpt-image-1.5— in · — out · — context
- gpt-image-1-mini— in · — out · — context
- gpt-image-1— in · — out · — context
- GPT-6.1 Sol$2 in · $10 out · 1.05M context
- GPT-6 Sol$2 in · $10 out · 1.05M context
- GPT-6 Luna$0.1 in · $0.5 out · 1.05M context
- GPT-6 Astra$10 in · $50 out · 1.05M context
- GPT-5.6 Terra$2 in · $12 out · 1.05M context
- GPT-5.6 Sol$4 in · $20 out · 1.05M context
- GPT-5.6 Luna$0.2 in · $1.2 out · 1.05M context
- GPT-5.6$4 in · $20 out · 1.05M context
Which models from other providers are similar to GPT-Realtime-2.1?
These are the closest in context window and input price, one per provider.
- Command A PlusCohere · $2.5 in · $10 out · 128K context
- Palmyra X4 (US)Amazon Bedrock · $2.5 in · $10 out · 123K context
- Qwen2.5-VL 72B InstructAlibaba Qwen · $2.294 in · $6.881 out · 131K context
- Claude Opus 4.5 (latest)Anthropic · $5 in · $25 out · 200K context
- GLM-5V-TurboZhipu GLM · $5 in · $22 out · 200K context