
Speechmatics
Speech-to-text API for batch and real-time transcription, with multilingual, medical and voice agent models, aimed at regulated industries.
Opens speechmatics.com in a new tab.
How the 7.0 was reached
- Quality of output5/10
Only vendor claims (350ms latency, case studies, Enhanced model accuracy); the GitHub academy repo (168 stars) is a sample repo, not independent evidence of transcription quality. Source: speechmatics.com
- Ease of use8/10
$100 free credit with no card, API key from portal, pip-installable SDKs, and a docs site with examples and a tutorial repo. Source: speechmatics.com
- Pricing value8/10
Published prices from $0.12/hr, $100 free credit, $25 credit packs with no commitment, and volume discounts up to 25%. Source: speechmatics.com
- API and integration quality8/10
Documented batch, realtime and search APIs, pip SDKs, and integrations with LiveKit, Pipecat, Vapi, Twilio and Zapier. Source: docs.speechmatics.com
- Solves the problem it claims to solve6/10
Compliance (HIPAA, GDPR, SOC 2, ISO 27001) and cloud/on-prem deployment match the regulated-industry claim, but this is vendor-stated; agent STT is cloud-only. Source: speechmatics.com
7.0/10 is the average of the five criterion scores above. The evidence comes from 3 different sources.Scores come from public evidence (the tool's own pages and documentation, and independent reviews); we do not test every tool hands-on. Scored 17 November 2026, on version 1 of our five-criteria scoring method.
Pricing
- New accounts get $100 in credit, no card required
- Access to the speech-to-text API for testing quality and language support
- Credit packs from $25 with no recurring commitment
- Annual subscriptions from $120/yr for 126 credits (5% off)
- Higher tiers up to $15,000/yr for 20,000 credits (25% off)
- Pre-recorded Melia 1 at $0.12/hr, real-time at $0.20/hr
- Custom enterprise pricing on every model
- Volume discount of up to 25% on credits
- Contact sales
For the full list of what each plan includes, and current prices, see speechmatics.com.
Key Features
- Pre-recorded and real-time speech-to-text through an API
- Melia 1 multilingual model that, per Speechmatics, transcribes 55+ languages and switches language mid-sentence
- Oak 1 medical model for drug names, dosages and procedure terms
- Linden 1 voice agent model with built-in turn detection for real-time use
- Integrations listed in the docs for LiveKit, Pipecat, Vapi, Twilio and Zapier
Pros & Cons
Pros
- Self-serve rates are published per hour for each model, so costs are easy to estimate
- $100 starting credit with no card lowers the cost of a trial
- Dedicated medical and voice agent models rather than one general model
- Docs list SDKs and integrations with common voice agent frameworks
Cons
- Rates vary by model and by batch versus real-time, so the price list takes some reading
- The Enhanced model costs $0.38 to $0.40/hr, well above Melia 1
- Accuracy and customer results on the site are the vendor's own claims and were not independently checked
- Real-time costs more than pre-recorded on the same model
Speechmatics is a developer-focused speech-to-text API with clear per-hour pricing and specialised models for medical and voice agent work. It suits teams that need multilingual or regulated-industry transcription. Compare accuracy on your own audio against Deepgram and AssemblyAI before committing.
Added to scored.tools on
For the Speechmatics team: put this score on your website
A small image (a badge) that shows this score and links back to this page. Free for any tool, at any score. It always shows the current score, so it changes when the score does. Copy this code into your website where you want the badge:
<a href="https://scored.tools/tools/speechmatics/"><img src="https://scored.tools/badge/speechmatics.svg" alt="Speechmatics on scored.tools: 7.0/10" width="160" height="40"></a>Do you work for Speechmatics? Claim this page: prove you work there, then confirm the facts, flag a mistake or add a short reply. It is free and never changes the score.
Competitors to Speechmatics
Other tools in the Voice category worth comparing.
Deepgram
7.4/10Unified Speech-to-Text, Text-to-Speech, and Voice Agent APIs built for enterprise scale and real-time accuracy.
AssemblyAI
7.3/10Speech-to-text APIs for pre-recorded and realtime audio in 99 languages, with speech understanding features on top.
Whisper
7.8/10OpenAI's open source speech model for multilingual transcription, translation and language identification.
Rev
6.4/10Transcription service with AI and human options, plus tools for case files, legal transcripts and document drafting.
More Articles Featuring Speechmatics
Speechmatics Review 2026: Pricing, Models and Limits
Speechmatics is a speech-to-text API with medical, multilingual and voice agent models and published hourly rates. Here's what the pricing and docs show.
reviewAssemblyAI Review 2026: Speech-to-Text API, Pricing, Limits
AssemblyAI covers transcription, streaming, speech understanding and voice agents from one API. Here is what the pricing and docs show, and where it falls short.