What is audio intelligence?

Audio intelligence is a layer of features that turn a transcript into structured, usable data: speaker labels, translations, entities, summaries, redacted PII, and more. With Gladia, these run on top of transcription in the same API call, so you go from raw audio to enriched output without chaining separate tools.

How is audio intelligence different from speech-to-text?

Speech-to-text produces the words; audio intelligence adds meaning on top of them. Transcription gives you the raw transcript, and audio intelligence features like summarization, entity detection, and PII redaction enrich that transcript in the same request.

Do audio intelligence features cost extra?

Some do. Speaker diarization, custom vocabulary, and code-switching are included in every request. Summarization (0.03€/hr), sentiment analysis (0.02€/hr), translation (0.05€/hr), named entity recognition and PII redaction (0.07€/hr each) are per-hour add-ons, and Audio-to-LLM is billed per token. You enable any feature by toggling a parameter, and pay only for the ones you turn on.

Does audio intelligence work in real time or only on pre-recorded audio?

Both. Gladia's async API supports the full feature set, while the real-time API covers a growing subset: custom vocabulary, named entity recognition, sentiment analysis, summarization, and chapterization (with post-session summaries returned by callback once the stream closes). A few features, including Audio-to-LLM and PII redaction, are currently pre-recorded only.

Which languages does audio intelligence support?

Gladia supports 100+ languages, with native code-switching when speakers change language mid-conversation. Audio intelligence features are built to be multilingual, so enrichment works across the same language set as transcription.

How does Gladia handle PII and compliance?

PII redaction replaces detected entities like names and phone numbers with labeled placeholders before data is stored. It's opt-in per request and available for pre-recorded transcription only. Gladia is GDPR, HIPAA, SOC 2 Type II, and ISO 27001 compliant with data residency options across EU and US.

Intelligence suite

Audio intelligence

Audio intelligence turns every conversation into structured data your product can use. Gladia returns transcription plus summaries, sentiment analysis, named entity recognition, and translation into 100+ languages, all in one API call. Built for developers of meeting assistants, contact centers, and voice agents.

Get your API key Get your API key

€50

Transcription credits

Test transcription and audio intelligence on your own audio. No expiry, no credit card.

API call

Transcript, insights, and translation together, with no extra pipeline to maintain.

Speaker diarization

3x fewer speaker errors than leading APIs, and the lowest error rate of any provider.

Trusted by over 350,000 users and 2,000+ enterprise teams

Do more with every audio file

Gladia layers enrichment features on top of transcription in one request: diarization, translation, entity recognition, Audio-to-LLM, PII redaction, and more. Accurate text is just the start.

Speaker diarization

Attribute speech to the right speaker with pyannoteAI Precision-2, the lowest diarization error rate of any STT provider in benchmarks.

Learn more

Audio-to-LLM

Send audio in and get summaries, action items, and custom outputs back from one request, choosing from 400+ LLMs to process it.

Learn more

Code-switching

Keep transcripts accurate when speakers switch languages mid-sentence, with automatic language detection that adapts in real time.

Learn more

Speech-to-text translation

Translate each utterance into any of 100+ languages, with the original timestamps and speaker labels kept on the translated output.

Learn more

Custom vocabulary

Teach the model your brand names, acronyms, and industry jargon with keyterm prompting, so specialized terms come out right.

Learn more

PII redaction

Mask names, emails, phone numbers, and other sensitive data in transcripts automatically, before it leaves your pipeline.

Learn more

Sentiment analysis

Classify each utterance as positive, negative, or neutral, with every result linked to its speaker and timestamp.

Learn more

Named entity recognition

Detect names, dates, organizations, and other entities in speech and return them as structured data.

Learn more

Summarization

Generate summaries of long conversations in concise, general, or bullet-point formats.

Learn more

See the full list See the full list

HOW IT WORKS

Add intelligence in one call

Every audio intelligence feature runs in the same API call as transcription. You add each one by toggling a single parameter, with no separate models, pipelines, or extra requests to manage.

Send your audio

Upload a file or stream in real time. One endpoint handles both async and live.

Turn on what you need

Enable any feature, from diarization to translation, with a single flag each. No extra pipelines to build.

Get structured results

Receive clean, enriched output that's labeled, translated, and redacted, ready to drop straight into your product.

Match the model to your workflow

Gladia offers two transcription models: Solaria-3 for the highest accuracy on European real-world audio, and Solaria-1 for the widest language coverage across any domain. Both support the full set of audio intelligence features.

Solaria-3

Highest accuracy on European real-world audio

#1 on European business audio and conversational speech
Optimized for EN, FR, DE, ES, IT
#1 on English production audio

Learn more

Solaria-1

Maximum language coverage across any domain

The most multilingual model: 100+ languages
3x lower diarization errors than other vendors
Code-switching, handled natively

Learn more

We are 100% benchmark and evaluation driven. Gladia was one of the best providers selected on merit to transcribe user videos, especially for non-English languages. Their reactive customer support and data compliance make their offer really compelling.

Core features included, add-ons when you need them

Speaker diarization, custom vocabulary, and 100+ languages are included at no extra cost. Add-ons like summarization, translation, sentiment analysis, named entity recognition, and PII redaction are billed only when you turn them on.

Self-serve

Flexible pay-as-you-go for moderate audio volumes

Starts at

0.16€/hour

Start with 50€ free credits

Enterprise

Annual plan with custom models, fine-tuning, and more

Custom

Top-up & save

Earn free credits when you top up

Free credits by top-up amount
Top up	Bonus	You get
1,000€	+10%	100€ in free credits
2,500€	+15%	375€ in free credits
5,000€	+20%	1,000€ in free credits

See our pricing

Plug in your key, start enriching

FAQs

Audio intelligence is a layer of features that turn a transcript into structured, usable data: speaker labels, translations, entities, summaries, redacted PII, and more. With Gladia, these run on top of transcription in the same API call, so you go from raw audio to enriched output without chaining separate tools.
Speech-to-text produces the words; audio intelligence adds meaning on top of them. Transcription gives you the raw transcript, and audio intelligence features like summarization, entity detection, and PII redaction enrich that transcript in the same request.
Some do. Speaker diarization, custom vocabulary, and code-switching are included in every request. Summarization (0.03€/hr), sentiment analysis (0.02€/hr), translation (0.05€/hr), named entity recognition and PII redaction (0.07€/hr each) are per-hour add-ons, and Audio-to-LLM is billed per token. You enable any feature by toggling a parameter, and pay only for the ones you turn on.
Both. Gladia's async API supports the full feature set, while the real-time API covers a growing subset: custom vocabulary, named entity recognition, sentiment analysis, summarization, and chapterization (with post-session summaries returned by callback once the stream closes). A few features, including Audio-to-LLM and PII redaction, are currently pre-recorded only. Since real-time coverage keeps expanding, check Gladia's live transcription docs for the latest.
Gladia supports 100+ languages, with native code-switching when speakers change language mid-conversation. Audio intelligence features are built to be multilingual, so enrichment works across the same language set as transcription.
PII redaction replaces detected entities like names and phone numbers with labeled placeholders before data is stored. It's opt-in per request and available for pre-recorded transcription only. Gladia is GDPR, HIPAA, SOC 2 Type II, and ISO 27001 compliant with data residency options across EU and US.