Speech & conversational AI
Amazon Nova vs Azure Speech / Foundry Agent Service
Amazon Nova Sonic (speech-to-speech on Bedrock) maps to Azure Voice Live and voice-based prompt agents in Foundry (preview), which also support realtime audio models.
Side-by-side comparison
| Aspect | Amazon Web Services | Microsoft Azure |
|---|---|---|
| Service | Amazon Nova | Azure Speech Foundry Agent Service |
| Lifecycle | Generally available | Custom neural voice and some avatar features are Limited Access. Voice-based agents and some monitoring/eval/telephony capabilities are in preview. |
| What it is | Amazon's first-party foundation models. Understanding (Nova Micro/Lite/Pro, Nova 2 Lite), creative (Nova Canvas image, Nova Reel video), speech-to-speech (Nova 2 Sonic), and multimodal embeddings. | Speech to text (real-time, fast, and batch transcription, custom speech), neural text to speech (custom voice), speech translation, pronunciation assessment, avatars, Voice Live, and edge containers. Managed agent platform. Prompt agents (declarative), voice-based prompt agents (Voice Live), and hosted agents (your container or zip, any framework), plus the Responses API for "ephemeral" agents. Shared toolbox via a managed MCP endpoint. |
| Pricing model | Per input and output speech/text token for Nova Sonic. | Voice Live usage plus model usage. |
| Docs | Amazon Nova docs | Azure Speech docs Foundry Agent Service docs |
| Pricing page | Amazon Nova pricing | Azure Speech pricing Foundry Agent Service pricing |
| Exams | Generative AI Developer – Professional | AI-103 |
Key differences
- Nova 2 Sonic is a single speech-to-speech foundation model you call through Bedrock's bidirectional streaming.
- Voice Live is a managed orchestration service (speech recognition, turn-taking, model interaction, synthesis) that works with realtime models or a cascaded STT → LLM → TTS chain. Foundry voice agents add telephony and avatar options (preview).
Pricing summary
Pricing models only. No prices are quoted because they change by region and over time.
- AWS
- Per input and output speech/text token for Nova Sonic. Amazon Nova pricing →
- Azure
- Voice Live usage plus model usage. Azure Speech pricing → Foundry Agent Service pricing →
Certification exam mapping
| Exam | Provider | Level | Status |
|---|---|---|---|
| AWS Certified Generative AI Developer – Professional | AWS | Professional | current Exam code not confirmed from the official page at build time; shown by name only. |
| Microsoft Certified – Azure AI Apps and Agents Developer Associate AI-103 | Azure | Associate | current Generative AI and agentic solutions with Microsoft Foundry using Python. Successor in scope to the retired AI-102. |
Relevance is based on each exam's published scope, not a line-by-line objective match. Check the official study guide.
FAQ
What is the Azure equivalent of Amazon Nova?
Amazon Nova Sonic (speech-to-speech on Bedrock) maps to Azure Voice Live and voice-based prompt agents in Foundry (preview), which also support realtime audio models.
How does pricing compare between Amazon Nova and Azure Speech and Foundry Agent Service?
AWS: Per input and output speech/text token for Nova Sonic. Azure: Voice Live usage plus model usage. Check the official pricing pages for current rates.
Which certification exams cover this topic?
AWS Certified Generative AI Developer – Professional; Microsoft Certified – Azure AI Apps and Agents Developer Associate (AI-103).