Skip to main content

​
Overview

If you’re using HyperWhisper in an environment with network filtering (corporate firewalls, parental controls, content filters, etc.), you’ll need to whitelist specific domains to ensure the app functions properly.
Important: You only need to whitelist the domains for services you actually use. For example, if you only use HyperWhisper Cloud for transcription, you don’t need to whitelist OpenAI, Groq, or other third-party providers.

​
Required Domains

These domains are always required for HyperWhisper to function properly, regardless of which transcription provider you use:
Without these domains whitelisted, core features like license validation and software updates will not work.
DomainPurpose
www.hyperwhisper.comMain website, license validation, API backend, and software updates
polar.shCustomer portal for license management
sentry.ioError reporting and crash analytics (optional but recommended)

​
HyperWhisper Cloud Endpoints

If you’re using HyperWhisper Cloud (the default, built-in transcription service):
  • https://transcribe-prod-v1.hyperwhisper.com/ - Transcription endpoint
  • https://transcribe-prod-v1.hyperwhisper.com/usage - Credit balance queries
HyperWhisper Cloud requires no API key and is the recommended option for most users. It uses a credit-based system with generous trial credits and affordable licensed pricing.

​
Model Downloads

If you plan to use local transcription with on-device Whisper models or local Gemma post-processing:
DomainPurpose
huggingface.coDownload Whisper.cpp and Gemma models
Specifically, HyperWhisper downloads models from:
  • https://huggingface.co/ggerganov/whisper.cpp/resolve/main/* - Whisper models
  • https://huggingface.co/unsloth/gemma-3-1b-it-GGUF/resolve/main/* - Gemma 1B models
  • https://huggingface.co/ggml-org/gemma-3-4b-it-GGUF/resolve/main/* - Gemma 4B models
  • https://huggingface.co/unsloth/gemma-3-12b-it-GGUF/resolve/main/* - Gemma 12B models

​
Optional: Third-Party Transcription Providers

Only whitelist the providers you use. If you’re not using a specific provider (e.g., you don’t have an OpenAI API key), you don’t need to whitelist their domains.

​
OpenAI Whisper

If you’ve configured OpenAI Whisper in Settings → API Keys:
DomainPurpose
api.openai.comWhisper transcription API and GPT post-processing
platform.openai.comAPI key management page (informational)
Specific endpoints:
  • https://api.openai.com/v1/audio/transcriptions - Whisper API
  • https://api.openai.com/v1/chat/completions - GPT post-processing
  • https://api.openai.com/v1/models - Model availability checks

​
Groq

If you’ve configured Groq in Settings → API Keys:
DomainPurpose
api.groq.comFast Whisper inference and language models
console.groq.comAPI key management page (informational)
Specific endpoints:
  • https://api.groq.com/openai/v1/audio/transcriptions - Whisper API
  • https://api.groq.com/openai/v1/models - Model availability checks

​
Fireworks AI

If you’ve configured Fireworks AI in Settings → API Keys:
DomainPurpose
api.fireworks.aiWhisper v3 and v3-turbo transcription
audio-prod.api.fireworks.aiWhisper v3 endpoint
audio-turbo.api.fireworks.aiWhisper v3-turbo endpoint
app.fireworks.aiAPI key management page (informational)
Specific endpoints:
  • https://audio-prod.api.fireworks.ai/v1/audio/transcriptions - Whisper v3
  • https://audio-turbo.api.fireworks.ai/v1/audio/transcriptions - Whisper v3-turbo
  • https://api.fireworks.ai/inference/v1/embeddings - Health check

​
Deepgram

If you’ve configured Deepgram in Settings → API Keys:
DomainPurpose
api.deepgram.comAdvanced speech-to-text API
deepgram.comWebsite (informational)
Specific endpoints:
  • https://api.deepgram.com/v1/listen - Transcription API
  • https://api.deepgram.com/v1/projects - Health check

​
AssemblyAI

If you’ve configured AssemblyAI in Settings → API Keys:
DomainPurpose
api.assemblyai.comAsync transcription API with advanced features
www.assemblyai.comWebsite (informational)
Specific endpoints:
  • https://api.assemblyai.com/v2/upload - Audio upload
  • https://api.assemblyai.com/v2/transcript - Transcription creation/polling

​
ElevenLabs Scribe

If you’ve configured ElevenLabs in Settings → API Keys:
DomainPurpose
api.elevenlabs.ioElevenLabs Scribe speech-to-text API
elevenlabs.ioWebsite and API key management (informational)
Specific endpoints:
  • https://api.elevenlabs.io/v1/speech-to-text - Scribe API
  • https://api.elevenlabs.io/v1/models - Model availability checks

​
Mistral Voxtral

If you’ve configured Mistral in Settings → API Keys:
DomainPurpose
api.mistral.aiVoxtral transcription API and model availability checks
console.mistral.aiAPI key management page (informational)
Specific endpoints:
  • https://api.mistral.ai/v1/audio/transcriptions - Voxtral Mini transcription API
  • https://api.mistral.ai/v1/models - Health check and model availability

​
Optional: AI Post-Processing Providers

If you’ve enabled AI post-processing to enhance transcriptions (fix typos, add punctuation, format text):
AI post-processing is optional. If you don’t use it, you don’t need to whitelist these domains.

​
Anthropic Claude

If you’ve configured Anthropic Claude for post-processing:
DomainPurpose
api.anthropic.comClaude AI models for text enhancement
console.anthropic.comAPI key management page (informational)
Specific endpoints:
  • https://api.anthropic.com/v1/chat/completions - Text post-processing
  • https://api.anthropic.com/v1/models - Model availability checks

​
Google Gemini

If you’ve configured Google Gemini for post-processing:
DomainPurpose
generativelanguage.googleapis.comGemini AI models for text enhancement
aistudio.google.comAPI key management page (informational)
ai.google.devGemma documentation (informational)
Specific endpoints:
  • https://generativelanguage.googleapis.com/v1beta/openai/chat/completions - Text post-processing
  • https://generativelanguage.googleapis.com/v1beta/models - Model availability checks

​
Local Gemma (Offline)

If you’re using local Gemma models for post-processing, you need to allow localhost connections:
DomainPurpose
127.0.0.1 or localhostLocal llama.cpp server for offline post-processing
Local Gemma runs entirely on your device and requires no internet connection. You only need to whitelist localhost if your firewall blocks local network connections.
Default port: 8080 (configurable in Settings → API Keys → Local Gemma) Specific endpoints:
  • http://127.0.0.1:8080/v1/chat/completions - Text post-processing
  • http://127.0.0.1:8080/health - Health check

​
Offline Mode

HyperWhisper can work completely offline if you use:
  1. Local transcription (libwhisper.cpp or Parakeet TDT v3):
    • Download Whisper models once from huggingface.co
    • Models are stored locally in ~/Library/Application Support/hyperwhisper/models/
    • No internet required after initial download
  2. No post-processing or local Gemma post-processing:
    • Local Gemma runs entirely on your device
    • No external API calls
  3. 7-day offline grace period for license validation:
    • HyperWhisper caches license validation for 24 hours
    • You can use the app offline for up to 7 days
    • After 7 days, you’ll need internet to revalidate
License validation still requires periodic internet access to hyperwhisper.com. The app will remind you to connect when the grace period expires.

​
Troubleshooting

​
Common Error Messages

These are actual error messages from HyperWhisper that indicate network filtering issues:
Error MessageLikely CauseSolution
”Network connection error”General connectivity issue or domain blockedCheck internet connection, verify required domains are whitelisted
”No internet connection”Not connected to internet or all domains blockedConnect to internet, whitelist required domains
”Cloud transcription requires an internet connection”Using cloud provider while offlineConnect to internet or switch to local transcription mode
”Network error: [details]“Specific network issue (timeout, DNS, etc.)Check domain whitelisting and network settings
”Cannot find host” / “Cannot connect to host”DNS blocked or domain not whitelistedWhitelist the specific domain (transcribe-prod-v1.hyperwhisper.com, api.openai.com, etc.)
”Unauthorized: invalid [Provider] API key”API key rejected (could be network blocking validation)Verify API key is correct, check if API domain is whitelisted
”Request timed out”Slow connection or partial blockingCheck connection speed, verify no SSL inspection interference
”[Provider] is unreachable”Provider domain blocked or offlineWhitelist provider’s API domain
”Server error (5xx)“Provider having issues OR SSL inspection breaking requestsCheck provider status, verify SSL bypass for provider domains
SSL inspection can cause many of these errors even when domains are whitelisted. If you see persistent “Network error” or “Unauthorized” messages despite correct API keys and whitelisting, see the SSL/TLS Inspection section below.

​
SSL/TLS Inspection Issues

Some corporate networks and content filters use SSL/TLS inspection (also called SSL bumping or HTTPS interception). This means:
  1. Your network filter intercepts HTTPS connections
  2. It presents its own certificate instead of the real one
  3. It decrypts, inspects, then re-encrypts your traffic
Even if you’ve whitelisted all the correct domains, SSL inspection can still cause connection failures because HyperWhisper validates SSL certificates for security.
If you’ve whitelisted all domains but still see “No network connection” or SSL certificate errors, SSL inspection is likely the cause.
Solution: Ask your IT administrator to add HyperWhisper domains to the SSL inspection bypass list (not just the URL allowlist). This ensures traffic to these domains passes through without certificate interception. Domains to bypass SSL inspection for:
  • *.hyperwhisper.com (includes transcribe-prod-v1.hyperwhisper.com)
  • polar.sh
  • Any third-party API domains you use (e.g., api.openai.com, huggingface.co)
Bypassing SSL inspection for HyperWhisper is actually better for your privacy—it means your voice transcriptions aren’t being inspected by your network filter.

​
How to Test Connectivity

If you’re experiencing issues, you can test connectivity to specific domains: 
# Test HyperWhisper Cloud
curl -I https://transcribe-prod-v1.hyperwhisper.com/

# Test license server
curl -I https://www.hyperwhisper.com/api/license/validate

# Test OpenAI (if using)
curl -I https://api.openai.com/v1/models

# Test Hugging Face (for model downloads)
curl -I https://huggingface.co
If curl commands succeed but HyperWhisper still fails, SSL inspection is likely interfering. The curl command uses your system’s certificate store, while HyperWhisper validates certificates more strictly.

​
Security & Privacy

​
Data Handling

  1. API Keys: Stored securely in macOS Keychain, never sent to HyperWhisper servers
  2. License Keys: Only sent to hyperwhisper.com for validation (encrypted via HTTPS)
  3. Audio Data:
    • HyperWhisper Cloud: Sent to transcribe-prod-v1.hyperwhisper.com (Cloudflare Workers edge network, processed in-memory, never stored on disk)
    • Third-party providers: Sent directly to their APIs (OpenAI, Groq, Deepgram, etc.) - subject to their privacy policies
    • Local models: Never leaves your device - completely offline and private

​
Network Security

  1. TLS/HTTPS: All network connections use encrypted HTTPS
  2. Certificate Validation: HyperWhisper validates SSL certificates to prevent man-in-the-middle attacks
  3. No Telemetry: HyperWhisper doesn’t track usage, analytics, or user behavior
  4. Crash Reports: Optional error reporting via Sentry (can be disabled in settings)

​
Privacy with Network Filters

When using corporate network filters with SSL inspection:
  • Without SSL bypass: Your network can decrypt and read all transcription content
  • With SSL bypass: Your transcriptions remain private and encrypted end-to-end
For maximum privacy, request IT to bypass SSL inspection for HyperWhisper domains.
Related Documentation: