Bring your own AI
Get an AI key and connect it in minutes
Interviewdom works with 11 AI providers for answers, plus dedicated speech-to-text options. You pay the provider directly for what you use, and your key stays encrypted on your PC.
Which provider should I pick?
Interviewdom needs an answer engine and a listening engine (speech-to-text). Some providers do both with one key. Not sure? Read the full comparison.
Cheapest
- Listening
- The same key
Free tier available and one key covers answers, screenshots and listening.
Best answers
- Listening
- Groq (GPT-OSS 20B)
Top-quality models for answers; Groq is a fast, low-cost listener.
Company AWS account
- Listening
- Groq (GPT-OSS 20B)
Answers are billed on your AWS account.
Private / local
- Listening
- Groq (GPT-OSS 20B)
Answers are generated on your own PC; only audio goes to Groq.
One key for everything
- Listening
- The same key
Answers, screenshots and listening with a single key.
Start here
First-time setup: from sign-up to your first mock interview
Sign up, start your trial, install Interviewdom on Windows, add your AI key, check your audio and run a first mock interview.
Easy~15 min
Which AI provider should I pick?
A quick guide to choosing an AI provider for Interviewdom: cheapest, best answers, company AWS, private/local, or one key for everything.
Easy~3 min
Speech-to-text: how Interviewdom listens
Why Interviewdom needs speech-to-text, which providers can do it, and how to add a second key just for listening.
Easy~4 min
Set spend limits so a long session can't run up a bill
How to cap or watch your spending at each AI provider you use with Interviewdom.
Easy~10 min
Troubleshooting AI provider errors in Interviewdom
Fix Test connection failures, key errors, rate limits, proxy/VPN problems and Windows audio issues.
Easy~5 min
Compare providers
Generated from the same provider catalog the app uses, so it always matches what you can pick in Settings → AI Providers.
| Provider | Answers | Reads screenshots | Speech-to-text | Recommended model | Cost estimate |
|---|---|---|---|---|---|
| OpenAI | Yes | Yes | Yes | GPT-6.1 Terra | Show estimateEstimate: about $0.14 per 1-hour interview for answers with GPT-6.1 Terra ($1 in / $5 out per 1M tokens), about $0.01 with GPT-6 Luna ($0.10 / $0.50), plus about $0.27 for 60 minutes of gpt-transcribe ($0.0045/min). Assumptions in costAssumptions; reasoning tokens are billed as output and are not included. |
| Anthropic (Claude) | Yes | Yes | No | Claude Haiku 5.5 | Show estimateEstimate: about $0.01 per 1-hour interview with Claude Haiku 5.5 ($0.10 in / $0.50 out per 1M tokens for prompts up to 100k tokens), about $0.28 with Sonnet 5.5 ($2 / $10), about $0.56 with Opus 5.5 ($4 / $20). Thinking tokens are billed as output and are not included. Speech-to-text is extra (separate provider). |
| AWS Bedrock | Yes | Yes | No | Claude Haiku 5.5 (global) | Show estimateEstimate: on-demand per-token pricing set by AWS; see https://aws.amazon.com/bedrock/pricing/. |
| Google Gemini (AI Studio) | Yes | Yes | Yes | Gemini 3.5 Flash-Lite | Show estimateEstimate: about $0.05 per 1-hour interview for answers with Gemini 3.5 Flash-Lite ($0.30 in / $2.50 out per 1M tokens) plus about $0.06 to transcribe 60 minutes of audio with the same model (115,200 audio tokens at $0.30/1M plus about 10,000 output tokens). Free tier available for Flash-Lite, Hydrogen and 3.8 Flash, but free-tier prompts may be used to improve Google products. |
| Azure OpenAI | Yes (your deployment) | Depends on deployment | Yes | Your deployment | Show estimateEstimate: depends on the model you deploy; see https://azure.microsoft.com/pricing/details/cognitive-services/openai-service/. |
| Groq | Yes | Some models (Qwen 3.8 27B (preview, reads screenshots)) | Yes | GPT-OSS 20B | Show estimateEstimate: speech-to-text about $0.04 per 1-hour interview with whisper-large-v3-turbo ($0.04/hour, 10 s minimum per request). |
| OpenRouter | Yes | Yes | Yes | Claude Haiku 5.5 via OpenRouter | Show estimateEstimate: depends on the model. With anthropic/claude-haiku-5.5 (from $0.10 in / $0.50 out per 1M tokens on OpenRouter) about $0.01 per 1-hour interview. OpenRouter shows a price range per provider for each model. |
| Mistral | Yes | Some models (Ministral 3 8B, Mistral Large 4) | Yes | Ministral 3 8B | Show estimateEstimate: about $0.01 per 1-hour interview with Ministral 3 8B ($0.15 per 1M tokens on its model page) and about $0.08 with Mistral Large 4 ($0.70 in / $2.10 out). |
| DeepSeek | Yes | Some models (DeepSeek V4.1 Flash) | No | DeepSeek V4.1 Flash | Show estimateEstimate: about $0.04 per 1-hour interview with deepseek-flash at peak prices ($0.30 per 1M input on cache miss, $1.20 per 1M output), about half that off-peak. Speech-to-text is extra (separate provider). |
| xAI (Grok) | Yes | Some models (Grok 4.7) | Not in Interviewdom yet | Grok 4.20 (non-reasoning) | Show estimateEstimate: about $0.13 per 1-hour interview with grok-4.20-0309-non-reasoning ($1.25 in / $2.50 out per 1M tokens), about $0.23 with grok-4.7 ($2 / $6). xAI speech-to-text $0.10/hour (REST) if the xai-native adapter is built. |
| Ollama (local, free) | Yes | Yes | No | Gemma 4 E4B (about 6.6–9.5 GB) | Show estimateEstimate: $0 in API fees; runs on your own PC (electricity only). Speed depends on your GPU/RAM. |
| Deepgram (speech-to-text only) | No (listening only) | No | Yes | – | Show estimateEstimate: about $0.26 per 1-hour interview with Nova-3 Monolingual pre-recorded ($0.0043/min), $0.31 with Nova-3 Multilingual ($0.0052/min). New accounts get $200 credit. |
How estimates are made: Estimate for a 1-hour live interview: about 40 AI answers, each about 2,000 input tokens (system prompt, resume, recent transcript) and 300 output tokens = about 80,000 input + 12,000 output tokens per hour. Speech-to-text: 60 minutes of audio. Screenshots, retries, reasoning/thinking tokens and long-context surcharges are NOT included and can raise the real cost. Prices are list prices read on verifiedAt. Each provider bills you directly; Interviewdom never adds a margin.
Guides by provider
OpenAI
Create an OpenAI API key, add credits, set a spend limit and connect it to Interviewdom. One key covers answers, screenshots and listening.
Easy~5 minCard needed
Anthropic (Claude)
Create a Claude API key in the Claude Console, buy credits, set a spend limit and connect it to Interviewdom. Add a second key for listening.
Easy~5 minCard needed
AWS Bedrock
Connect Interviewdom to Claude or Amazon Nova on AWS Bedrock with a Bedrock API key or an IAM access key. Includes the least-privilege policy and model access steps.
Advanced~15 minCard needed
Google Gemini (AI Studio)
Create a free Gemini API key in Google AI Studio and connect it to Interviewdom. One key covers answers, screenshots and listening.
Easy~5 minNo card needed
Azure OpenAI
Create an Azure OpenAI resource, deploy a model, copy the endpoint and key, and connect them to Interviewdom.
Advanced~20 minCard needed
Groq
Create a Groq API key for fast, cheap speech-to-text (Whisper Large v3 Turbo) and fast open models for answers.
Easy~5 minNo card needed
OpenRouter
One OpenRouter key gives you access to many AI models. Buy credits, create a key, pick a model ID and connect it to Interviewdom.
Easy~5 minCard needed
Mistral
Create a Mistral API key, set up billing and connect it to Interviewdom. Ministral 3 and Mistral Large 4 can read screenshots.
Easy~5 minCard needed
DeepSeek
Top up DeepSeek, create an API key and connect it to Interviewdom. DeepSeek Flash can now read screenshots; add a second key for listening.
Easy~5 minCard needed
xAI (Grok)
Load xAI credits, create a Grok API key and connect it to Interviewdom. Grok 4.7 can read screenshots.
Easy~5 minCard needed
Ollama (local, free)
Install Ollama on Windows, download a Gemma 4 model and connect it to Interviewdom. No key, no API fees; answers stay on your PC.
Medium~20 minNo card needed
Deepgram (speech-to-text only)
Deepgram is a speech-to-text-only provider. New accounts get $200 free credit with no card. Create a key and use it for listening.
Easy~5 minNo card needed
Your experience got you the interview. Say it clearly.
30 minutes free · bring your own AI key · Windows 10 & 11