Use AgentRouter to transcribe audio, generate speech, and inspect speech model supply through one routed voice API for agents.
2
Live Providers
5
Live Routes
10 credits ($0.01)
Transcribe
10 credits ($0.01)
Speak
The live speech catalog currently exposes 3 capabilities for AI agents. Transcribe currently starts at 10 credits ($0.01), and speak currently starts at 10 credits ($0.01).
Availability Note
Deepgram and OpenAI transcribe and speak routes are live. `speech.analyze.deepgram.mpp` and `speech.speak.grok.mpp` are still temporarily unavailable.
What Is Speech & Voice?
Agent speech is the operational layer that lets an AI agent turn audio into text, turn text into speech, and inspect the live speech model catalog programmatically.
Speech workflows often split across different providers for transcription, speech generation, and model listing. That fragmentation makes it harder to keep routing, pricing, and fallback logic stable inside a production agent.
AgentRouter keeps those tasks behind one routed speech domain so the agent can compare live routes before execution and switch providers without rewriting its speech workflow.
Top Scenarios
Recover text from an audio clip when the agent needs speech-to-text before downstream reasoning.
Deepgram
MPP
10 credits ($0.01)
OpenAI
MPP
10 credits ($0.01)
Meeting notes, voice agents, and transcription workflows
Coverage: 1 capabilities · 2 live routes
Start with Speech Transcribe
Create spoken output when the workflow starts from text and needs a voice response.
Deepgram
MPP
10 credits ($0.01)
OpenAI
MPP
10 credits ($0.01)
Voice replies, assistants, and narration workflows
Coverage: 1 capabilities · 2 live routes
Start with Speech Synthesize
List model or voice supply before locking the workflow to one route.
Speech Models List
Deepgram
0 credits ($0)
Route selection, quality checks, and voice catalog inspection
Coverage: 1 capabilities · 1 live routes
Start with Speech Models List
Provider Comparison
The rows below are derived from the live route catalog currently exposed through AgentRouter.
| Provider | Best For | Key Capabilities | Starting Price | Available Via |
|---|---|---|---|---|
D Deepgram Covers Speech Transcribe, Speech Synthesize, Speech Models List through the live routed catalog. 3 live routes | Speech Transcribe and Speech Synthesize | Speech Transcribe, Speech Synthesize, Speech Models List Modes: Managed | 0 credits ($0) | MPP |
O OpenAI Covers Speech Transcribe, Speech Synthesize through the live routed catalog. 2 live routes | Speech Transcribe and Speech Synthesize | Speech Transcribe, Speech Synthesize Modes: Managed |
Quick Start
Get an API key, ask AgentRouter to recommend the best route for the job, and execute through one shared wallet instead of wiring each provider separately.
Step 1
Enable the API once
Create one AgentRouter API key that can be reused across all live domains and capabilities.
Step 2
Recommend the route
Let AgentRouter compare the current routes, prices, and route availability before the first execute call.
Step 3
Execute the capability
Run the selected workflow through the chosen route while AgentRouter handles pricing, wallet debits, and upstream settlement.
Works with
Pricing By Task
These examples translate the current live route prices into workflow questions operators usually evaluate before shipping a domain into an agent.
| Task | Route | Price | Notes |
|---|---|---|---|
| Transcribe one audio clip | Deepgram (speech.transcribe.deepgram.mpp) | 10 credits ($0.01) | Return normalized text from speech input. |
| Generate one speech response | Deepgram (speech.speak.deepgram.mpp) | 10 credits ($0.01) | Create audio output from a text prompt. |
| Inspect speech model supply | Deepgram (speech.models.list.deepgram.mpp) | 0 credits ($0) | Check live provider catalog before hard-coding a route. |
Pricing note: 1000 credits = $1 USD. Raw API responses still return numeric credit fields such as creditsCharged.
Why AgentRouter
Recommendation happens before execution, so the agent can compare current route supply and pricing instead of hard-coding one provider forever. The same routed contract stays stable even as provider coverage or transport paths change underneath.
FAQ
It is the routed layer that lets an AI agent transcribe audio, generate speech, and inspect speech model supply through one voice domain.
Use transcribe when the input is audio and the output should be text. Use speak when the input is text and the output should be generated audio.
Models-list lets the agent inspect the live provider catalog before choosing a transcription route or voice surface, which reduces hard-coded assumptions.
No. AgentRouter keeps the live speech routes behind one wallet and one routed API surface.
Reference
Use the capability browser below when you want contract-level detail: endpoints, live example modals, route breakdowns, and exact route prices.
| Capability | Endpoints | Price |
|---|---|---|
transcribe Convert spoken audio into text. | GET /api/agentic-api/domains/speech/capabilities/transcribeRead the capability contract, current lifecycle status, and machine-readable metadata for this speech workflow. | Free |
POST /api/agentic-api/domains/speech/capabilities/transcribe/recommendAsk AgentRouter to compare eligible routes inside this capability and return the best recommendation before execution. | Free | |
POST /api/agentic-api/domains/speech/capabilities/transcribe/executeExecute this capability through the selected route while AgentRouter handles pricing, wallet debits, and upstream settlement. | ||
speak Convert text into generated speech. | GET /api/agentic-api/domains/speech/capabilities/speakRead the capability contract, current lifecycle status, and machine-readable metadata for this speech workflow. | Free |
POST /api/agentic-api/domains/speech/capabilities/speak/recommendAsk AgentRouter to compare eligible routes inside this capability and return the best recommendation before execution. | Free | |
POST /api/agentic-api/domains/speech/capabilities/speak/executeExecute this capability through the selected route while AgentRouter handles pricing, wallet debits, and upstream settlement. 1 route in this breakdown are temporarily unavailable. | ||
analyze Analyze spoken or transcribed content for higher-level signals. | GET /api/agentic-api/domains/speech/capabilities/analyzeRead the capability contract, current lifecycle status, and machine-readable metadata for this speech workflow. | Free |
POST /api/agentic-api/domains/speech/capabilities/analyze/recommendAsk AgentRouter to compare eligible routes inside this capability and return the best recommendation before execution. | Free | |
POST /api/agentic-api/domains/speech/capabilities/analyze/executeExecute this capability through the selected route while AgentRouter handles pricing, wallet debits, and upstream settlement. 1 route in this breakdown are temporarily unavailable. | ||
models.list List provider speech models or voices when the upstream route explicitly supports it. | GET /api/agentic-api/domains/speech/capabilities/models-listRead the capability contract, current lifecycle status, and machine-readable metadata for this speech workflow. | Free |
POST /api/agentic-api/domains/speech/capabilities/models-list/recommendAsk AgentRouter to compare eligible routes inside this capability and return the best recommendation before execution. | Free | |
POST /api/agentic-api/domains/speech/capabilities/models-list/executeExecute this capability through the selected route while AgentRouter handles pricing, wallet debits, and upstream settlement. |
Related Pages
AgentRouter landing page
Browse all live domains, top API cards, and route coverage across the platform.
Install and auth
Enable AgentRouter, create the API key, and connect the shared wallet flow.
Phone, SMS & Voice domain
Create managed phone agents, provision numbers, inspect conversations and calls, and keep telephony workflows behind one supplier-neutral phone domain.
Web Crawl & Extraction domain
Fetch, scrape, extract, crawl, map, browse, and screenshot public pages through one routed surface.
10 credits ($0.01)
| MPP |