AssemblyAI
assemblyai.com
Integrate speech-to-text, speaker labels and audio intelligence models through a developer API. Usage is billed per audio hour after promotional credits.
Product at a glance
- Category
- Transcription
- Pricing type
- Paid
- Free plan
- No
- Listed platforms
- Web
- Processing
- Cloud
- Commercial use
- Unknown
- API access
- Available
- Pricing evidence
- 2026-11-18
What to know about AssemblyAI
Integrate speech-to-text, speaker labels and audio intelligence models through a developer API. Usage is billed per audio hour after promotional credits.
Who it may suit: Engineering teams adding asynchronous or streaming transcription, speaker labels and PII redaction via API.
Technical details, licensing and privacy notes
- Capabilities
- Transcription · Speech to text · AI audio
- Languages
- 99 languages on Universal speech-to-text models
- Input formats
- Not verified
- Output formats
- Not verified
- Free allowance
- Not verified
- Maximum file size
- Not verified
- Maximum duration
- Not verified
- Usage limits
- Not verified
- Privacy summary
- Not verified
- Facts reviewed
- Not yet reviewed
A quick decision guide
Before you choose
Who this tool may suit
Engineering teams adding asynchronous or streaming transcription, speaker labels and PII redaction via API.
What to confirm
Confirm current plan limits, commercial-use rights and available features with the provider before committing.
Based on public directory information and editorial prompts, not a hands-on product test.
How to evaluate this tool
Editorial checks to perform with your own sample, not hands-on test results.
- 1
Calculate total per-hour cost when combining base transcription with diarization or summarization add-ons.
- 2
Test custom vocabulary boosting and speaker labeling on multi-party recordings.
- 3
Review regional endpoints, BAAs or data-retention settings before sending regulated audio.
Plans and evidence
Pricing and availability can change. Review the source before using paid features.
Pricing & plans
Pricing model
Paid
AssemblyAI bills speech-to-text and audio intelligence models per audio hour on a usage-based meter. One-time promotional API credits are offered for evaluation, not an ongoing recurring free plan.
Published plan details
Usage-based API
Usage-basedPer-second metered transcription; speaker diarization and audio intelligence add-ons billed separately.
Data completeness
Coverage of key fields backed by dated official sources. This is a documentation checklist, not a product quality score.
3/10
verified fields
Still to verify: Platforms, Processing mode, Languages, File formats, Usage limits, Commercial use, Privacy. Missing data does not mean the product lacks the feature.
Last source dates are shown in the Data sources & confidence section below.
Data sources & confidence
We distinguish publisher documentation from submitted claims and editorial catalog labels. An unknown field means it has not been verified, not that the product lacks the feature.
- API availabilityOfficial source
API-first speech-to-text, diarization and audio intelligence platform.
Checked 2026-11-18View source - Ongoing free planOfficial source
Usage-based per-hour billing after one-time promotional evaluation credits; not an ongoing free plan.
Checked 2026-11-18View source - Ongoing free planOfficial source
Ongoing free-plan status is classified from the dated publisher pricing evidence. Trial-only access is not counted as an ongoing free plan.
Checked 2026-11-18View source - PlatformsEditorial metadata
Platform labels are curated from directory metadata and have not all been independently checked against current publisher compatibility documentation.
- PricingOfficial source
AssemblyAI bills speech-to-text and audio intelligence models per audio hour on a usage-based meter. One-time promotional API credits are offered for evaluation, not an ongoing recurring free plan.
Checked 2026-11-18View source - PricingOfficial source
Per-second metered transcription; speaker diarization and audio intelligence add-ons billed separately.
Checked 2026-11-18View source
Explore workflows and comparisons
Related guides
Transcription FAQ
Are AI transcripts always accurate?
No. Accuracy varies with audio quality, accents, language, overlap and vocabulary. Review important passages against the original recording before relying on a transcript.
Should I upload confidential meeting audio?
Only after checking your organization policy, participant permissions and the provider terms for storage, access, training use, sharing and deletion.