All tasks / Voice & audio / Voice agents
Best AI APIs for voice agents
Models you can call through a paid API, ranked from public benchmarks, with their price in this task’s unit.
STop ranked: Gemini 3.8 Live (Google), 1,088 on Artificial Analysis Speech Agent Arena (models).6 options listed · data verified 18 Oct 2026
Ranked list
| Tier | Name | Approach | Runs | Rated by | Price | Skill | Compare |
|---|---|---|---|---|---|---|---|
| S |  Gemini 3.8 LiveGoogle | Models | Cloud | 1,088 on Artificial Analysis Speech Agent Arena (models) | $0.01/min of audio | — | |
| S |  Grok Voice 3.0xAI | Models | Cloud | 1,070 on Artificial Analysis Speech Agent Arena (models) | $0.08/min of audio | — | vs Gemini 3.8 Live |
| A |  GPT-Live 1OpenAI | Models | Cloud | 1,057 on Artificial Analysis Speech Agent Arena (models) | $0.05/min of audio | — | vs Gemini 3.8 Live |
| B | GPT-Realtime-2.1 OpenAI | Models | Cloud | 941 on Artificial Analysis Speech Agent Arena (models) | $0.19/min of audio | — | vs Gemini 3.8 Live |
| C | GPT-Realtime-2.1 Mini OpenAI | Models | Cloud | 844 on Artificial Analysis Speech Agent Arena (models) | $0.08/min of audio | — | |
| C |  Qwen Audio 3.0 Realtime PlusAlibaba (Qwen) | Models | Cloud | 792 on Artificial Analysis Speech Agent Arena (models) | $0.07/min of audio | — |
Tiers come only from public benchmarks (Artificial Analysis Speech Agent Arena (models)); tools without a result of their own inherit the tier of the best model they offer. How we rank.