EvoHire MCP server and API
Run AI technical interviews from Claude, ChatGPT or your own agent
The EvoHire MCP server lets Claude, ChatGPT and other MCP-compatible assistants use EvoHire as a tool: create interview jobs, invite candidates, and read scored results with integrity verdicts. Building your own pipeline? Use the EvoHire API with an API key.
Included on every plan for a limited time. 14 days of every Professional feature, no credit card required.
Last updated: 26 October 2026
Which candidates finished their interview this week? Rank them by score.
Three candidates finished the Senior Backend Engineer interview:
- 1PNPriya N.ExcellentLow concern
- 2DMDaniel M.GoodLow concern
- 3SKSam K.GoodLikely AI-assisted
Sam K. is flagged. Want the evidence from the report?
What is the EvoHire MCP server?
MCP (Model Context Protocol) is an open standard that lets AI assistants use tools in other software. EvoHire’s MCP server connects your assistant to your EvoHire account, so you can run first-round technical interviews by asking for them in plain language instead of clicking through the app.
| Works with | Claude and ChatGPT, plus other assistants that support remote MCP servers |
|---|---|
| Server URL | https://api.evohire.ai/mcp |
| Sign-in | OAuth. Your assistant asks you to log in to EvoHire, with nothing to configure |
| Hosting | Hosted by EvoHire. Nothing to install |
| Tools | 50+ tools for jobs, candidates, invites and interview results |
| Plans | Every plan, including Free, for a limited time |
| Usage | No separate limit. Interviews use your interview credits, the same as in the dashboard |
What your agent can do with EvoHire
The EvoHire MCP server has more than 50 tools covering the whole first-round screening workflow. Your agent works with the same access as the EvoHire account that signed in.
| Area | What your agent can do |
|---|---|
| Jobs and interviews | Create interview jobs from a job description, set screening questions, review and edit interview questions, publish or close a job |
| Candidates and invites | Add candidates, send interview invites, and check who is invited, in progress or finished |
| Results | Read each candidate’s scores against your rubric, the interview summary, and the integrity verdict with its evidence |
| Pipeline reporting | Get counts and status across all your open roles |
Need a full reference? See the EvoHire Help Center.
Set up in a few minutes
You need an EvoHire account, on any plan. Your assistant handles the rest. When you connect, it asks you to sign in to EvoHire.
https://api.evohire.ai/mcpClaude
- In Claude, open Settings, then Connectors.
- Choose to add a custom connector.
- Name it EvoHire and paste https://api.evohire.ai/mcp as the server URL.
- Select Connect and sign in to EvoHire when prompted.
- In a new chat, turn on the EvoHire connector and ask Claude to list your open roles.
ChatGPT
- In ChatGPT, open Settings, then the section for apps and connectors.
- If adding a custom connector requires it, turn on developer mode in that section’s advanced settings.
- Create a new connector, name it EvoHire, and paste https://api.evohire.ai/mcp as the server URL, with OAuth as the authentication.
- Sign in to EvoHire when prompted.
- In a new chat, select the EvoHire connector and ask ChatGPT to list your open roles.
Other MCP clients
Any client that supports remote MCP servers with OAuth can connect. Add https://api.evohire.ai/mcp as a remote server and sign in to EvoHire when prompted.
Using the API instead
- In the EvoHire app, go to User Settings > API Keys and create a key.
- Treat the key as a secret: it has access to your EvoHire account. Store it in your secrets manager, never in code.
- Revoke it in the same place at any time.
Step-by-step guides are in the EvoHire Help Center.
Example workflows
Copy a prompt, paste it into Claude or ChatGPT, and change the details.
Create a role
Here’s the job description for a Senior Backend Engineer (Go, PostgreSQL). Create an EvoHire interview for it and show me the questions before you publish.Invite a shortlist
Invite these candidates to the Senior Backend Engineer interview: [paste names and emails].Review results
Which candidates finished their interview this week? Rank them by score, and list anyone with a Likely AI-assisted or High risk verdict separately.Pipeline check
For each of my open roles, how many candidates are invited, in progress and finished?Agency client summary
Write a client-ready summary of the top 3 candidates for the Data Engineer role, with their scores and integrity verdicts.
Who it’s for
Recruitment and staffing agencies
Screen candidates for every client from the assistant you already use, and turn results into client-ready shortlists.
EvoHire for agenciesIn-house talent teams
Keep hiring managers updated and move candidates forward without leaving your chat.
EvoHire for talent teamsTeams building their own hiring agents
Add a technical interviewer with cheat detection to your agent through MCP or the API.
See setup stepsMCP server or API?
| MCP server | API | |
|---|---|---|
| Best for | Working with EvoHire from Claude, ChatGPT and other assistants | Custom pipelines, scheduled jobs and your own agent code |
| Setup | Add the server URL and sign in to EvoHire | Create a key in User Settings > API Keys |
| Who drives it | You, in plain language | Your code |
| Usage | Interview credits, same as the dashboard | Interview credits, same as the dashboard |
Security and access
- Your assistant connects with OAuth and acts with the same access as your EvoHire account. MCP doesn’t use an API key: you sign in to EvoHire the usual way, with a code sent to your email.
- Destructive actions and some other write actions ask for your confirmation in Claude or ChatGPT before they run.
- API keys give access to your EvoHire account. Create and revoke them in User Settings > API Keys.
- Integrity verdicts reach your agent exactly as they appear in the EvoHire report.
- When you connect EvoHire to an AI assistant, the data your agent reads is processed by that assistant’s provider under your agreement with them.
Frequently asked questions
What is an MCP server?
MCP (Model Context Protocol) is an open standard that lets AI assistants use tools in other software. EvoHire’s MCP server lets your assistant create interviews, invite candidates and read results in your EvoHire account.
Which AI assistants work with EvoHire?
Claude and ChatGPT are tested and supported. Other assistants that support remote MCP servers with OAuth sign-in can connect with the same server URL.
Do I need a paid plan?
No. MCP and API access is included on every plan, including Free, for a limited time.
Do interviews run through MCP cost extra?
No. There’s no separate MCP or API limit. Interviews use your interview credits exactly as they do in the dashboard.
Is there an API as well?
Yes. Create an API key in User Settings > API Keys to call EvoHire from your own code or agent.
Can my agent change or delete things without asking?
Destructive actions and some other write actions ask you to confirm in your assistant before they run.
Can my agent see how cheat detection works?
No. Your agent gets the same integrity verdict and evidence you see in the EvoHire report. Answers are analysed after the interview, inside EvoHire.
Does EvoHire integrate with my ATS?
EvoHire connects to Greenhouse, Lever, Ashby and other ATSs inside the app. See all integrations.
How do I disconnect?
Remove the EvoHire connector in your assistant’s settings. For API keys, revoke the key in User Settings > API Keys.
Give your recruiting agent a technical interviewer
Connect EvoHire in minutes and screen every technical candidate from the assistant you already use.
14 days of every Professional feature. No credit card required.