# Deepgram vs Kaldi

**Leader by Vioscale score:** Deepgram

| Attribute | Deepgram | Kaldi |
|---|---|---|
| **Vioscale score** | 68.3 (43% (low)) | 17.8 (15% (low)) |
| activity.commits_last_30d | - | 0 |
| adoption.dependent_repos | - | 1 |
| adoption.github_stars | - | 15,469 |
| deployment.options | `{"cloud":true,"self_hosted":true}` | `{"cloud":true,"self_hosted":true}` |
| description.long | A platform providing real-time speech-to-text, text-to-speech, and voice agent APIs with native support for understanding turn-taking, interruptions, and multi-speaker scenarios. Supports 45+ languages with features like speaker identification, smart formatting, and keyword recognition for enterprise and developer use cases. | A free, open-source framework for building and deploying speech recognition systems, supporting multiple platforms from desktop to mobile and in-browser execution. |
| integrations.count | 7 | - |
| integrations.list | `[{"name":"Zapier"},{"name":"Genesys"},{"name":"Cisco"},{"name":"Avaya"},{"name":"Nuance"},{"name":"UniMRCP"},{"name":"Y Combinator"}]` | - |
| language.primary | - | Shell |
| market.availability | `{"primaryMarkets":["US","GB","EU"],"availabilityScope":"global","availableCountries":[],"notAvailableCountries":[]}` | - |
| platform.support | `{"cli":true,"web":true}` | `{"cli":true,"mac":true,"web":true,"linux":true,"android":true,"windows":true}` |
| pricing | `{"type":"usage","plans":[{"free":true,"name":"Pay As You Go","summary":"From $0.0048/min STT, $0.015/1k chars TTS. $200 free credit.","features":["Speech-to-text (45+ languages)","Text-to-speech","Voice Agent API","Speaker diarization","Smart formatting","Entity detection","Automatic language detection","Keyterm prompting","PII redaction"],"commitment":"monthly","components":[{"per":{"qty":1,"unit":"minute"},"kind":"metered","amount":0.0048,"currency":"USD","description":"Nova-3 Monolingual STT (baseline)"},{"per":{"qty":1,"unit":"minute"},"kind":"metered","amount":0.0065,"currency":"USD","description":"Flux English STT"},{"per":{"qty":1000,"unit":"characters"},"kind":"metered","amount":0.015,"currency":"USD","description":"Aura-1 TTS"},{"per":{"qty":1,"unit":"minute"},"kind":"metered","amount":0.056,"currency":"USD","description":"Voice Agent API Standard tier"}],"description":"Metered usage with initial $200 free credit","contactSales":false,"includedLimits":{"initial_credit":"$200"}},{"free":false,"name":"Enterprise","summary":"Custom pricing. Contact sales.","features":["Custom trained models","On-premises or private cloud deployment","Dedicated support","HIPAA Business Associate Agreement","Data Processing Agreement"],"description":"Custom solutions for high volumes, specialized deployment, or compliance requirements","contactSales":true}],"addOns":[{"name":"Redaction","components":[{"per":{"qty":1,"unit":"minute"},"kind":"metered","amount":0.002,"currency":"USD"}]},{"name":"Keyterm Prompting","components":[{"per":{"qty":1,"unit":"minute"},"kind":"metered","amount":0.0013,"currency":"USD"}]},{"name":"Entity Detection"},{"name":"Speaker Diarization"}],"summary":"From $0.0048/min for speech-to-text, $0.015/1k characters for text-to-speech. Free tier with $200 credit.","currency":"USD","freeTier":true,"sourceUrl":"https://deepgram.com/pricing","retrievedAt":"2026-08-19T21:33:14.342Z","billingPeriods":["month"]}` | `{"type":"open_source","freeTier":true,"sourceUrl":"https://kaldi-asr.org/","retrievedAt":"2026-08-20T12:17:47.249Z"}` |
| pricing.free_tier | yes | yes |
| pricing.model | freemium | open_source |
| pricing.price_level | low | free |
| pricing.transparent | yes | - |
| security.gdpr | yes | - |
| security.hipaa | yes | - |
| security.pci | yes | - |
| security.scorecard | - | 3.9 |
| security.soc2 | yes | - |
| security.vulnerabilities | - | `{"count":0,"source":"https://advisories.ecosyste.ms/api/v1/advisories?ecosystem=conda&package_name=kaldi&per_page=100","last_12m":0,"max_severity":null}` |

## Capabilities (Speech To Text)

| Capability | Deepgram | Kaldi |
|---|:--:|:--:|
| **Capabilities** |  |  |
| Real time streaming | - | - |
| Speaker diarization | - | - |
| Custom vocabulary | - | - |
| Fine tuning | - | - |
| Word timestamps | - | - |
| Punctuation formatting | - | - |
| Pii redaction | - | - |
| Vertical model | - | - |
| Offline on device | - | - |
| Open source | - | - |

*Source: Vioscale. Generated 2026-09-01T17:19:39.416Z. "-" = undocumented, not absent.*