Cookie consent notice

By clicking “Accept All”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.

Accept all

Deny

Preferences
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Build With Velma

The Velma API offers industry-leading models for transcription, deepfake detection, emotion detection, and more, at 10x lower cost.

Talk to Sales
Self-Serve
Deepfake Detection
Only $0.25 / hour
Industry-leading accuracy
Batch & Streaming APIs
1,000 free credits
Self-Serve
Transcription
Starting at $0.03 / hour
World-class accuracy on real-world data
Batch & Streaming APIs
400 hours in free credits
Custom API Access
Talk to sales
Cutting-edge models at enterprise scale
Priority access to new endpoints and models
Bulk usage discounts
Dedicated support options available
Self-Serve
Deepfake Detection
Only $0.25 / hour
Industry-leading accuracy
Batch & Streaming APIs
1,000 free credits
Self-Serve
Transcription
Starting at $0.03 / hour
World-class accuracy on real-world data
Batch & Streaming APIs
400 hours in free credits
Custom API access
Talk to Sales
Cutting-edge models at enterprise scale
Priority access to new endpoints and models
Bulk usage discounts
Dedicated support options available
Deepfake Detection
Transcription

Deepfake Detection Pricing Breakdown

Model
Included Features
Credit Rate
Dollar Rate
Velma Deepfake Detect Batch
Use the world’s leading synthetic voice detector to prevent fraud or deception.
Segment-based probability scores, every four seconds with two second overlap. Only 3 seconds minimum required.
25/hr
$0.25/hr
Velma Deepfake Detect Streaming
Use the world’s leading synthetic voice detector to prevent fraud or deception.
Segment-based probability scores, every four seconds with two second overlap. Only 3 seconds minimum required.
25/hr
$0.25/hr

See How We Compare

Feature
Modulate
Resemble AI
Reality Defender
Sensity AI
Top Model
Velma-2 Deepfake Detect (Synthetic) 
Detect-3B Omni
Real Suite
Sensity Deepfake Detection Hub
Pricing
$0.25 /hr
$144/hr, with max discounts down to $28/hr
Not listed
Not listed
Self-Serve API
Yes
Yes
Yes
No, talk to sales
Accuracy (higher is better)
Unavailable
Equal Error Rate (lower is better)
Unavailable
Unavailable
Audio Required for Result
3 seconds
Not listed
6 seconds
30 seconds
Optimized for…
• Enterprise fraud prevention & financial security
• Contact center routing and authentication
• Media/journalism disinformation detection
• Adversarial deepfakes generated by expert actors
• Enterprise fraud prevention & financial security
• Multimodal (audio + video + image) threat detection
• Video conferencing user verification
• Brand & executive protection
• Media/journalism disinformation detection
• Law enforcement & judicial/forensic investigations
• Government & defense intelligence
Additional Models Available
Transcription, Emotion, PII Redaction, Conversation Analytics
Speaker Identification + Voice Biometrics
Video & Image Deepfake Detection
Court-admissible forensic reports with chain-of-custody logs and NIST/ENFSI-aligned documentation
Free Credits
Up to 40 hrs
None
50 audio scans per month
None (must contact sales)

Speech-to-Text Pricing Breakdown

Model
Included Features
Dollar Rate
Credit Rate
Velma Transcribe Batch
Using Dynamic Ensemble Block technology, leverage multiple models seamlessly to get top-quality results.
Diarization, Emotion Detection, Accent Identification, PII/PHI Tagging, Multilingual Support
$0.03/hr
3/hr
Velma Transcribe Streaming
Using Dynamic Ensemble Block technology, leverage multiple models seamlessly to get top-quality results.
Diarization, Emotion Detection, Accent Identification, PII/PHI Tagging, Multilingual Support
$0.06/hr
6/hr
Velma Transcribe Batch English Fast
Directly access our fastest submodel for lightning-quick results.
Realtime Factor of up to 200x, English-Only
$0.025/hr
2.5/hr

See How We Compare

Feature
Modulate
Deepgram
AssemblyAI
ElevenLabs
Batch API Cost (Multilingual)
$0.03 / hr
$0.55 / hr
$0.21 / hr
$0.39 / hr
Streaming API Cost (Multilingual)
$0.06 / hr
$0.55 / hr
$0.45 / hr
$0.39 / hr
Earnings-22 Benchmark
7.6% WER
28.1% WER
24.7% WER
9.7% WER
Real-World Accuracy (AMI Corpus)
14.9% WER
28.1% WER
24.7% WER
26.2% WER
Speaker Diarization
Free
$0.12 / hr
$0.02 / hr
n/a
Emotion Detection
Free, 20+ emotions
Pay for sentiment
Sentiment - $0.02 / hr
n/a
Accent Identification
Free, 20+ accents
n/a
n/a
n/a
Multilingual Support
57 languages
51 languages
99 languages
90 languages
PII / PHI Tagging
Included
$0.12/hr
$0.08/hr
Unavailable
Credits Included
Up to 400 hrs
Up to 360 hrs
Up to 300 hrs
2.5 hrs

FAQ

What is a credit?

All API operations cost credits, with one credit equal to $0.01 (aside from volume discounts.) Different API operations consume credits at different rates.

Can I split credits between the API and Enterprise Platform?

Yes. Credits purchased, and volume discounts offered, through the Enterprise Platform can be extended to the API as well.Do you offer volume discounts?

Do you offer volume discounts?

Yes. Larger platform tiers include bigger credit bundles, and API pricing supports volume-based discounts. Talk to Sales to learn more.

Can we start small and scale?

Yes. Most customers start by paying as they go and expand as usage grows. Upgrading to a paid tier with a volume discount is quick and painless.

Get pricing tailored to your conversation volume

Talk with our team to estimate usage, compare platform tiers, and design a plan that fits your workflows and scale.