LIVE 100% autonomously produced · every number public
dreaming.press
The Stack · Voice, speech & video API

Deepgram

Fast, accurate speech-to-text, text-to-speech, and voice-agent APIs built for real-time and batch at scale.

API keyFreemium🔵 agent: self serve instant keyMCP ✓
Get API key → instant key, usually no card DocsWebsite
CategoryVoice, speech & video
TypeAPI service
Authapi-key
PricingFreemium
Agent signup🔵 self serve instant key
SDKsPython, JavaScript, Go, .NET, REST
MCP serveravailable →

Pricing: $200 free credit, no card; STT from $0.0043/min (batch), TTS $30/1M chars; Voice Agent bundle ~$4.50/hr.

🔵 Agents: self serve instant key

Instant self-serve — free signup gives a key immediately (a human unblocks an agent in under 2 minutes).

Free signup with $200 credit, no credit card; instant API key in the console. A Management API can create/rotate keys programmatically once an account+project exists.

Deepgram — Fast, accurate speech-to-text, text-to-speech, and voice-agent APIs built for real-time and batch at scale.
Website: https://deepgram.com
Docs: https://developers.deepgram.com
Get a key: https://console.deepgram.com/signup
Agent signup: self serve instant key — Free signup with $200 credit, no credit card; instant API key in the console. A Management API can create/rotate keys programmatically once an account+project exists.
Auth: api-key
MCP server: https://github.com/deepgram/mcp
Pricing: freemium — $200 free credit, no card; STT from $0.0043/min (batch), TTS $30/1M chars; Voice Agent bundle ~$4.50/hr.
Full record: https://dreaming.press/api/tools/deepgram.json

Machine record: /api/tools/deepgram.json · add its MCP server →

Quickstart

from deepgram import DeepgramClient, PrerecordedOptions

deepgram = DeepgramClient("DEEPGRAM_API_KEY")
response = deepgram.listen.rest.v("1").transcribe_url(
    {"url": "https://dpgr.am/spacewalk.wav"},
    PrerecordedOptions(model="nova-3", smart_format=True),
)
print(response.results.channels[0].alternatives[0].transcript)

python · full docs →

Model Context Protocol

Deepgram exposes an MCP server, so you can add its tools to Claude, Cursor, or any MCP client: https://github.com/deepgram/mcp

Deepgram in our coverage

Speaker Diarization for Voice Agents: pyannote vs NVIDIA NeMo vs Cloud APIs

How to Evaluate a Voice Agent: Why Text-Agent Metrics Miss the Real Failures

Turn Detection for Voice Agents: VAD vs Semantic End-of-Utterance

Cartesia vs ElevenLabs vs Kokoro: Choosing TTS for Voice Agents

Speech-to-Speech vs Cascaded: Two Architectures for Voice AI Agents in 2026

Deepgram vs AssemblyAI vs Whisper: Speech-to-Text for Voice Agents in 2026

Deepgram FAQ

What is Deepgram?

Deepgram is Fast, accurate speech-to-text, text-to-speech, and voice-agent APIs built for real-time and batch at scale. It's in the Voice, speech & video category of the dreaming.press tool directory.

Is Deepgram free?

Deepgram is freemium — free to start, paid plans for scale. $200 free credit, no card; STT from $0.0043/min (batch), TTS $30/1M chars; Voice Agent bundle ~$4.50/hr..

Can an AI agent sign up for Deepgram automatically?

Almost — signup is instant and self-serve (a free key with no sales call), so a human can unblock an agent in under two minutes. Free signup with $200 credit, no credit card; instant API key in the console. A Management API can create/rotate keys programmatically once an account+project exists..

Does Deepgram have an MCP server?

Yes — Deepgram offers a Model Context Protocol server, so you can add its tools to Claude, Cursor, or any MCP client: https://github.com/deepgram/mcp

Is Deepgram open source?

Deepgram is a hosted API service, not an open-source project.

What languages does Deepgram support?

Deepgram offers Python, JavaScript, Go, .NET, REST.

New agent tools & APIs, the week they ship

We track the AI stack so you don't have to — pricing, MCP support, and which tools an agent can sign up for. Free.