# Speech and voice AI for AI agents

As of 1 Nov 2026, an AI agent gets in on its own at the one product we drove in speech and voice AI.

Our drives of 1 speech and voice AI product, 1 Nov 2026.

**For an agent today:** [Deepgram](https://agentdrives.com/tools/deepgram?via=md), at L3 Does the work. MCP server: `https://api.dx.deepgram.com/kapa/mcp`.

## 1 product by level

1 of 1 let our agent in on its own.

- L5 Autonomous: 0
- L4 Runs on its own: 0
- L3 Does the work: 1 (Deepgram)
- L2 Gets in: 0
- L1 Needs a person: 0
- L0 Read-only: 0

## Every product we drove

Open a product for its whole drive.

Sorted by level, then work done, then time to the first call ([how we order lists](https://agentdrives.com/method?via=md#order)).

| Product | Level | Stopped at | Route | Core work | Driven |
|---|---|---|---|---|---|
| [Deepgram](https://agentdrives.com/tools/deepgram?via=md) | L3 Does the work | Paying: Not tried: paying needs an account | API, Browser | 5 of 9 | 1 Nov 2026 |

## Where agents drop out

No checkpoint stopped our agent here.

Of the 1 products driven, how many our agent passed on its own at each checkpoint.

| Checkpoint | Passed on its own | A person did it | Stopped | Nothing to do | Not reached | Not tried |
|---|---|---|---|---|---|---|
| Docs | 1 of 1 | 0 | 0 | 0 | 0 | 0 |
| Sign up | 0 of 1 | 0 | 0 | 1 | 0 | 0 |
| Verify | 0 of 1 | 0 | 0 | 1 | 0 | 0 |
| Key | 0 of 1 | 0 | 0 | 1 | 0 | 0 |
| First call | 1 of 1 | 0 | 0 | 0 | 0 | 0 |
| First action | 1 of 1 | 0 | 0 | 0 | 0 | 0 |
| Pay | 0 of 1 | 0 | 0 | 0 | 0 | 1 |
| Cancel | 0 of 1 | 0 | 0 | 1 | 0 | 0 |
| Delete | 0 of 1 | 0 | 0 | 1 | 0 | 0 |

## What an agent can do in each

The tasks most speech and voice AI products share. Our agent did at least one at 1 of 1.

| Product | Choose the AI model | Record and transcribe a meeting |
|---|---|---|
| Deepgram | done | blocked |

## The walls our agent met

A bot check stopped our agent at 1 of 1. At 1 of them it got in another way.

- Google reCAPTCHA (at Sign up): stopped 1 of 1 (Deepgram: got in another way)

## Use the data

The numbers on this page, for your own tools and agents.

- JSON: [/categories/speech.json](https://agentdrives.com/categories/speech.json?via=md)
- This page: [/categories/speech](https://agentdrives.com/categories/speech?via=md)

## Get your shortlist driven before you choose

A Tool comparison drives the tools you're choosing between and shows which ones your agents can use. $490 for three tools, ready in 48 hours.

[See the Tool comparison](https://agentdrives.com/stack-check) · [Check your own product free](https://agentdrives.com/scan?via=md)