Deepgram

Name: Deepgram
Brand: Deepgram
Rating: 4.6 (327 reviews)

Deepgram is a developer-focused voice AI platform that provides speech-to-text (real-time streaming and batch), text-to-speech, a unified Voice Agent API for speech-to-speech orchestration, and audio-intelligence features such as diarization, summarization, sentiment, and intent extraction.

Est 2015|Valuation $1.3B|Raised $246.4M|123 Ref Customers

LinkedIn 43K|Web traffic 579K|FTEs 375|HQ San Francisco, CA

MarketsAI Voice Agent Platform

01Competitive Landscape

Narrative analysis: where alternatives can excel

AssemblyAI vs. Deepgram - AssemblyAI emphasizes concurrency options and may present different packaging for enterprise buyers. Deepgram documents explicit concurrency limits on certain tiers which buyers should consider when evaluating real-time scaling. Speechmatics vs. Deepgram - Speechmatics emphasizes a broader set of supported languages in public materials; organizations with broad language needs should confirm coverage and evaluate accuracy per language. Google Cloud Speech-to-Text vs. Deepgram - Large cloud providers publish extensive pricing and enterprise discounts at scale and provide deep integration with other cloud services; Deepgram differentiates on voice-specific infrastructure and self-hosted voice AI options. Market share and adoption trends - Press reporting often cites vendor-reported customer counts; such figures should be treated as vendor-provided and may not constitute an audited market-share measure.

	Deepgram	AssemblyAI	Speechmatics
Starting Price (public)	Published examples show usage-based per-minute STT pricing and per-character TTS pricing; Pay-as-you-go options include introductory credit (vendor published examples).	Vendor-published STT pricing examples and model-specific tiers; alternative packaging for concurrency is emphasized.	Vendor-published pricing examples and free-tier minutes noted in public materials.
Best For	Low-latency real-time voice applications; voice agents; organizations seeking self-hosted options	Teams seeking pay-as-you-go models with higher concurrency options and broad feature set	Organizations prioritizing broad language coverage and enterprise deployment options
Key Strength	Unified Voice Agent API concept and self-hosted deployment option; developer-oriented APIs	Concurrency framing and broad feature set; pricing variations across models	Language breadth and deployment flexibility
Key Weakness	Reported concerns regarding cost scaling, concurrency limits on lower tiers, and batch endpoint latency in some conditions	Pricing and feature packaging vary by model; enterprise terms may require negotiation	Some enterprise pricing and feature granularity require sales engagement

01Competitive Landscape

Narrative analysis: where alternatives can excel

AssemblyAI vs. Deepgram

AssemblyAI emphasizes concurrency options and may present different packaging for enterprise buyers. Deepgram documents explicit concurrency limits on certain tiers which buyers should consider when evaluating real-time scaling.

Speechmatics vs. Deepgram

Speechmatics emphasizes a broader set of supported languages in public materials; organizations with broad language needs should confirm coverage and evaluate accuracy per language.

Google Cloud Speech-to-Text vs. Deepgram

Large cloud providers publish extensive pricing and enterprise discounts at scale and provide deep integration with other cloud services; Deepgram differentiates on voice-specific infrastructure and self-hosted voice AI options.

Market share and adoption trends

Press reporting often cites vendor-reported customer counts; such figures should be treated as vendor-provided and may not constitute an audited market-share measure.

	Deepgram	AssemblyAI	Speechmatics
Starting Price (public)	Published examples show usage-based per-minute STT pricing and per-character TTS pricing; Pay-as-you-go options include introductory credit (vendor published examples).	Vendor-published STT pricing examples and model-specific tiers; alternative packaging for concurrency is emphasized.	Vendor-published pricing examples and free-tier minutes noted in public materials.
Best For	Low-latency real-time voice applications; voice agents; organizations seeking self-hosted options	Teams seeking pay-as-you-go models with higher concurrency options and broad feature set	Organizations prioritizing broad language coverage and enterprise deployment options
Key Strength	Unified Voice Agent API concept and self-hosted deployment option; developer-oriented APIs	Concurrency framing and broad feature set; pricing variations across models	Language breadth and deployment flexibility
Key Weakness	Reported concerns regarding cost scaling, concurrency limits on lower tiers, and batch endpoint latency in some conditions	Pricing and feature packaging vary by model; enterprise terms may require negotiation	Some enterprise pricing and feature granularity require sales engagement