AI Rate CardGPT-Realtime vs GPT-Realtime-2.1: API price and specs
GPT-Realtime-2.1 is 90% cheaper than GPT-Realtime on input price.
GPT-Realtime-2.1 is 70% cheaper than GPT-Realtime on output price.
GPT-Realtime-2.1 is 84% cheaper than GPT-Realtime on a 1M-in / 200k-out workload.
GPT-Realtime-2.1 has a 300% larger context than GPT-Realtime.
| Field | GPT-Realtime | GPT-Realtime-2.1 |
|---|---|---|
| Lowest input, per 1M | $32.00 +900% | $3.20 |
| Median input, per 1M | $32.00 | $32.00 |
| Lowest output, per 1M | $64.00 +233.3% | $19.20 |
| Median output, per 1M | $64.00 | $64.00 |
| Lowest unit price | — | — |
| Cost of 1M in + 200k out | $44.80 +536.4% | $7.04 |
| Context | 32K +75% | 128K |
| Max output | 4K +87.2% | 32K |
| Providers | 2 +60% | 5 |
| Regions | — | — |
| Accepts | audio, image, text | audio, image, text |
| Returns | audio, text | audio, text |
| Type | Speech to speech | Speech to speech |
| Free option | no | no |
| Cheapest provider | fastrouter | cometapi |
Your own workload: how many tokens the model gets (input) and gives back (output) - for example, in a month. The "Cost" row of the table above is recalculated for these numbers, for each model.
Computed from current offers each time the page opens; the best value of a row is bold, the percent is how far behind the best.
Providers selling both
Frequently asked questions
Which is cheaper, GPT-Realtime or GPT-Realtime-2.1?
GPT-Realtime-2.1 is 90% cheaper than GPT-Realtime on input price. GPT-Realtime-2.1 is 70% cheaper than GPT-Realtime on output price. GPT-Realtime-2.1 is 84% cheaper than GPT-Realtime on a 1M-in / 200k-out workload.
Which has the larger context, GPT-Realtime or GPT-Realtime-2.1?
GPT-Realtime: 32K tokens; GPT-Realtime-2.1: 128K tokens.
Where can I use both GPT-Realtime and GPT-Realtime-2.1?
fastrouter, openai