Amazon · AI model
Amazon Nova 2.5 Sonic
Successor to Nova 2 Sonic for real-time voice agents, generally available on Amazon Bedrock since October 5, 2026, with the same API surface, voices and language support as Nova 2 Sonic and the same list prices in US East (N. Virginia).
Last verified Available 4 sources
Key specs
- Provider
- Amazon
- Family
- Amazon Nova
- API identifier
- amazon.nova-2-5-sonic
- Availability
- Available
- Release date
- 2026-10-05
- License
- Proprietary (AWS Customer Agreement)
- Context window
- 256,000 tokens
- Max output
- 64,000 tokens
- API access
- Yes
- Consumer app
- Unknown
Capabilities
Inputs
- Text input Yes Cross-modal input: speech and text in the same session
- Image input No
- Audio input Yes Speech input; the Bedrock model card lists 'Speech' as supported and generic 'Audio' as not supported
- Video input No
- File input Unknown
Outputs
- Text output Yes
- Image output No
- Audio output Yes Speech output
Features
- Reasoning Unknown
- Tool calling Yes Tool use and asynchronous task completion
- Web search Unknown
"Unknown" means we have not verified this capability yet; it does not mean the model lacks it.
Pricing
Estimate your monthly cost →| Dimension | Tier | Price | Since | Verified | Source |
|---|---|---|---|---|---|
| Input tokens | base | $0.33 per 1M tokens | 2026-10-12 | 2026-10-25 | Amazon Bedrock Pricing US East (N. Virginia), on-demand list price; speech-to-speech model pricing table. Text input tokens. Pricing page note: text token pricing applies to uses such as speech-to-text transcription, tool calls and adding conversation history to the session. |
| Output tokens | base | $2.75 per 1M tokens | 2026-10-12 | 2026-10-25 | Amazon Bedrock Pricing US East (N. Virginia), on-demand list price; speech-to-speech model pricing table. Text output tokens. Pricing page note: text token pricing applies to uses such as speech-to-text transcription, tool calls and adding conversation history to the session. |
| Audio input tokens | base | $3.00 per 1M tokens | 2026-10-12 | 2026-10-25 | Amazon Bedrock Pricing US East (N. Virginia), on-demand list price; speech-to-speech model pricing table. Speech input tokens. |
| Audio output tokens | base | $12.00 per 1M tokens | 2026-10-12 | 2026-10-25 | Amazon Bedrock Pricing US East (N. Virginia), on-demand list price; speech-to-speech model pricing table. Speech output tokens. |
Prices are shown in the currency the provider publishes; we never convert currencies.
Compare Amazon Nova 2.5 Sonic
Guides
About Amazon Nova 2.5 Sonic
Nova 2.5 Sonic is a speech-to-speech model (model ID amazon.nova-2-5-sonic) that AWS announced as generally available on October 5, 2026. AWS states that it improves reasoning, instruction following and tool-calling accuracy and lowers latency compared with Nova 2 Sonic, while keeping the same API surface, voices and language support. It accepts speech and text input, produces speech and text output and supports asynchronous tool calling. The Bedrock model card lists a 256K-token context window, 64K max output tokens, the Standard service tier only and In-Region inference only. It is available in US East (N. Virginia), US West (Oregon), Europe (Stockholm) and Asia Pacific (Tokyo); the announcement states the same pricing as Nova 2 Sonic, and the pricing page lists identical rates in US East (N. Virginia).
Strengths and weaknesses
Strengths
- Same API surface, voices and language support as Nova 2 Sonic; identical list prices in US East (N. Virginia)
- Speech and text input in the same session; tool use with asynchronous tool calling
- AWS reports improved reasoning, instruction following and tool-calling accuracy over Nova 2 Sonic
Weaknesses
- Cannot be fine-tuned; behavior is customized only through system prompts and configuration parameters
- Does not support real-time speech-to-speech translation
- Officially supported for seven languages only
- 256K-token context window according to the Bedrock model card, smaller than the 1M tokens listed for Nova 2 Sonic
Sources
Last verified Page updated 2026-10-27