Gladia

Gladia is an audio API for teams building transcription and analysis into products, alongside a separate desktop dictation app. It is strongest for developers who need both live and uploaded-audio workflows; its usage-based tiers make volume and concurrency central to the choice.

Overview

Gladia handles asynchronous transcription and real-time speech, including live streaming, then adds tools such as speaker diarization, sentiment analysis, PII redaction, entity detection and audio-to-LLM processing. That breadth can turn a transcript into structured input for downstream workflows, rather than leaving teams to assemble separate analysis steps. The API-first product is less suited to people who simply want a ready-made meeting-notes application.

Key features

Two transcription modes: Async processing covers recordings, while real-time streaming serves live speech. Starter includes automatic language detection and switching across 100+ languages, useful for multilingual applications.

Audio intelligence: Speaker labeling, sentiment analysis, entity detection and PII redaction extend the output beyond plain text. This is valuable when an application needs speaker-aware or privacy-conscious processing, but the feature set is most relevant to teams able to build around an API.

Developer connections: Python and Node.js SDKs, WebSocket streaming and REST uploads provide multiple integration routes. Named integrations include Pipecat, LiveKit, Twilio and Retell; Zapier, Make and n8n connectors support workflow automation.

Security and compliance: Gladia says customer data is not used for model training by default and describes encryption in transit and at rest. It lists GDPR, HIPAA, SOC 2 Type II and ISO 27001 credentials, and says EU and US workloads use separate regional infrastructure. These controls matter for sensitive workloads, though customers should match them to their own requirements.

Desktop dictation: GladiaFlow provides real-time voice dictation on Mac and Windows. Its code is published on GitHub under the MIT license, a separate option from the core API service.

Pricing

Gladia has a free Starter tier and pay-as-you-go rates, with lower Growth rates tied to upfront commitment. Hourly call and total transcribed-audio limits vary by tier, so the headline rate alone does not establish whether a workload will fit.

PlanPrice and termsWhat it suitsStarterfree; pay-as-you-go, with async at $0.61/hr and real-time at $0.75/hr. Includes 30 real-time concurrent requests, 25 async concurrent requests and 50€ in free credits.A low-commitment way to begin or handle modest usage. New users' 50€ credits are one-time, do not reset monthly and do not expire; the concurrency ceilings may constrain heavier workloads.Starter - asynchronous$0.61 a month, billed Flexible pay-as-you-go; 30 real-time concurrent requests and 25 async concurrent requests.Recording-focused workloads that want flexible billing rather than an upfront commitment.Starter - real-time$0.75 a month, billed Flexible pay-as-you-go; 30 real-time concurrent requests and 25 async concurrent requests.Live transcription workloads that need the flexible Starter structure.GrowthCustom pricing; async as low as $0.20/hr and real-time as low as $0.25/hr, with upfront commitment. Flexible concurrent request limits, custom volume discounts and automatic model training opt-out.Higher-volume teams that can commit upfront to pursue lower rates and need more flexible concurrency. The stated rates are floors, not guaranteed prices for every workload.Growth - asynchronous$0.20 a month, billed As low as $0.20/hr; commit upfront to unlock savings.Async-heavy workloads willing to commit for the lower starting rate.Growth - real-time$0.25 a month, billed As low as $0.25/hr; commit upfront to unlock savings.Live workloads willing to commit for the lower starting rate.EnterpriseCustom pricing. Unlimited concurrent requests, 100+ languages, zero data retention, custom hosting, SLAs and limitless scaling.Organizations needing contractual service commitments, custom hosting or zero data retention, and capacity beyond Starter's concurrency limits.Starter and Growth include help center and Discord support; Enterprise adds premium support through dedicated Slack and an account manager. Starter's free credits have no monthly reset, so they are better treated as an initial usage balance than a recurring free allowance.

Platforms

The service is available on API and web, while GladiaFlow runs on macOS and Windows. That split makes Gladia suitable both for developers embedding audio functions and for desktop users who want voice dictation, but the desktop app is distinct from the API offering.

Who it's for

Gladia is a strong fit for developers and product teams building multilingual speech features, especially when they need live streaming, asynchronous processing and audio analysis in one service. Teams handling sensitive audio may value its stated security controls and compliance credentials. It is a weaker fit for individuals seeking a conventional meeting assistant or for workloads that cannot justify usage-based billing or Growth's upfront commitment.

Pros and cons

Pro: Live and async transcription, 100+ languages and audio-analysis capabilities cover several stages of an audio workflow.

Pro: SDKs, streaming and REST options, plus named integrations and automation connectors, give developers multiple connection paths.

Pro: Starter includes 50€ in non-expiring credits without a monthly reset, making it possible to begin without a recurring free quota.

Con: Starter concurrency is capped at 30 real-time and 25 async requests, which can limit simultaneous workloads.

Con: Growth's lower hourly rates require an upfront commitment, so they are less attractive when volume is uncertain.

Con: The developer-centered service is not a substitute for a dedicated meeting-notes product for users who do not need API integration.

Alternatives

Choose Happy Scribe if a free tier with unlimited meeting recordings and 45-minute recordings better fits your use. Grain is another option for meeting recording, with a free Notetaker seat and 45-minute meetings. Pick Notta for a free allowance of one seat and 120 transcription minutes per month. Trint is a paid alternative with a free trial. Speechmatics offers a free tier with $100 in credits and support for concurrent real-time sessions. For a free, open-source desktop option, noScribe runs on Linux, macOS and Windows. Otter.ai has a free plan with monthly transcription limits. Rev offers a free plan with English-only transcription and monthly AI transcription and caption minutes.

For broader comparisons, see Audio Transcription Software, Speech-to-Text Software, Speech Recognition Software and Transcription Software.

Verdict

Choose Gladia if you are building a speech-enabled product and want live and asynchronous transcription, multilingual coverage and audio analysis through developer interfaces. Its main advantage is breadth across the audio workflow; its main drawback is that price and capacity depend on usage, and the lower Growth rates require commitment. If you need a standalone meeting-notes tool rather than audio infrastructure, look elsewhere.