Get a Mistral API key for Interviewdom
Create a Mistral API key, set up billing and connect it to Interviewdom. Ministral 3 and Mistral Large 4 can read screenshots.
- Updated November 19, 2026
- Easy
- About 5 min
- Payment card needed
- Can also listen (speech-to-text)
1. At a glance
- What you get: answers, screenshot reading (Ministral 3 and Mistral Large 4), and speech-to-text with Voxtral.
- Free mode: Mistral API keys work in Free mode by default, with usage and rate limits meant for testing.
- Cost: pay as you go; about $0.01 per 1-hour interview with Ministral 3 8B (see section 8).
- Time needed: about 5 minutes.
2. Create an account
- Sign up for Mistral AI and create your organization.
3. Billing and a spend limit
- Set up billing for your organization. Mistral says API keys stay inactive until a payment method is on file (outside Free mode). It can take a few minutes after adding billing before keys work.
- Spend limit (recommended): Mistral lets you set spending limits at the organization level (not per workspace). Workspaces have no monthly cap by default.
4. Create the API key
- Open the Admin Panel → API → API Keys page.
- Select Create new key.
- Name it Interviewdom – laptop and choose the workspace.
- Copy the key now.
5. Pick a model
| Model | Best for | Reads screenshots |
|---|---|---|
Ministral 3 8B (ministral-8b-2512) – recommended | Fast, cheap | Yes |
Mistral Small 4 (mistral-small-2603) | Balanced | being verified |
Mistral Large 4 (mistral-large-4-0) | Best answers | Yes |
6. Connect in Interviewdom
- Go to Settings → AI Providers → Mistral.
- Paste your key.
- Choose a model (Ministral 3 8B is pre-selected).
- Click Test connection. You should see something like OK · 450 ms.
Your key is stored encrypted on your PC and is sent only to Mistral, never to Interviewdom.
7. Speech-to-text
Mistral can transcribe with Voxtral Mini Transcribe (model voxtral-mini-latest) through its audio transcriptions endpoint.
For now we recommend a Groq key for listening.
8. Costs in practice (estimate)
Assumptions: about 80,000 input + 12,000 output tokens per 1-hour interview.
- Ministral 3 8B ($0.15 per 1M tokens on its model page): about $0.01.
- Mistral Large 4 ($0.70 in / $2.10 out per 1M tokens): about $0.08.
Add your listening cost.
9. Troubleshooting
| Problem | What it means | Fix |
|---|---|---|
| 401 unauthorized | Key is wrong, deleted, or still inactive because billing isn't set up | Add a payment method, wait a few minutes, or create a new key. |
| 404 model not found | Model ID is wrong or retired | Pick a model from the list. |
| 429 rate limit | Free-mode or plan limits reached, or spending limit hit | Wait, set up billing, or raise the limit. |
| Screenshot not understood | The model can't read images | Use Ministral 3 or Mistral Large 4. |
| Network error | Proxy/VPN blocks api.mistral.ai | Ask IT to allow it. |
10. Security tips
- Never share your key, including with Interviewdom support.
- Delete a leaked key in the Admin Panel and create a new one.
- Interviewdom stores your key encrypted on your PC and sends it only to Mistral.