While FloWords excels at local transcription, you can optionally use cloud providers for specific needs. Cloud providers require your own API keys and send audio to external servers.
Faster Processing
Cloud servers can transcribe faster than some local setups, especially on older Macs.
Latest Models
Access the newest AI models without downloading large files.
Backup Option
Use as a fallback when local transcription encounters issues.
Specialized Features
Some providers offer unique features like speaker diarization.
The gold standard for cloud transcription.
Features :
High accuracy across languages
Consistent, reliable performance
Same Whisper technology as local models
Timestamp support
Pricing : ~$0.006 per minute of audio
Create account at platform.openai.com
Generate an API key
In FloWords: Settings > Transcription > Cloud Providers
Select OpenAI
Paste your API key
Test the connection
Ultra-fast transcription with Groq’s specialized hardware.
Features :
Extremely fast processing
Competitive accuracy
Cost-effective pricing
Low latency
Pricing : Pay per usage (check groq.com for current rates)
Create account at console.groq.com
Generate an API key
In FloWords: Settings > Transcription > Cloud Providers
Select Groq
Paste your API key
Test the connection
Real-time streaming transcription specialist.
Features :
Real-time streaming support
Speaker diarization
Custom vocabulary
Multiple languages
Pricing : Tiered pricing based on usage
Create account at deepgram.com
Generate an API key
In FloWords: Settings > Transcription > Cloud Providers
Select Deepgram
Paste your API key
Test the connection
Google’s multimodal AI with transcription capabilities.
Features :
Multimodal understanding
Context awareness
Advanced language support
Integration with Google ecosystem
Pricing : Check Google Cloud pricing
Set up Google Cloud account
Enable Gemini API
Generate credentials
In FloWords: Settings > Transcription > Cloud Providers
Select Gemini
Configure your credentials
Known for voice synthesis, also offers speech recognition.
Features :
High-quality speech recognition
Voice analysis capabilities
Good for voice-centric workflows
Pricing : Check elevenlabs.io for current rates
European AI provider with privacy focus.
Features :
European data processing
Strong multilingual support
GDPR compliance
Competitive pricing
Pricing : Check mistral.ai for rates
Provider Speed Accuracy Price Privacy Languages OpenAI Fast Excellent $$$ US servers 99+ Groq Fastest Great $$ US servers 50+ Deepgram Fast Great $$ US servers 30+ Gemini Fast Excellent $$$ Google Cloud 40+ ElevenLabs Medium Good $$ US servers 20+ Mistral Medium Great $$ EU servers 20+
Create Account - Sign up with your chosen provider
Generate API Key - Usually in Settings or API section
Copy Key - Keep it secure, don’t share
Configure FloWords - Add key in Settings > Transcription
Test Connection - Use the test button to verify
Select as Active - Choose as your transcription method
Never share your API keys
Use separate keys for different applications
Set spending limits where possible
Rotate keys periodically
Monitor usage for unexpected charges
FloWords can automatically fall back to cloud providers when local transcription fails.
Go to Settings > Transcription
Enable Fallback Mode
Set your primary method (usually local)
Configure cloud provider as fallback
Set fallback conditions (timeout, error, quality threshold)
Primary : Your selected local model
Secondary : Configured cloud provider
Tertiary : Apple’s native Speech Recognition
Check provider dashboards regularly
Set up billing alerts
Review monthly statements
Use local models when possible
Set cloud as fallback only
Choose cost-effective providers (Groq, Deepgram)
Batch longer recordings rather than many short ones
Use Case Recommendation Daily dictation Local models (free) Professional transcription Cloud provider Occasional backup Cloud fallback High volume Local + selective cloud
Double-check the key for typos
Ensure you copied the full key
Check if the key is still active
Generate a new key if needed
Wait before making more requests
Check your provider’s rate limits
Consider upgrading your plan
Use local models as primary
Check your internet connection
Verify the provider’s service status
Try a different provider
Fall back to local models
Review your usage history
Check for compromised API keys
Set spending limits
Contact provider support