
Argent for Claude Desktop on third-party inference
Claude Desktop's gateway mode points at Argent, so every chat is attributed, checked and signed. Here is what Argent governs, how it connects and what your team sets.
What Argent does with Claude Desktop on third-party inference
Every call Claude Desktop on third-party inference makes to a model goes through Argent first. Argent knows which agent sent it and which person owns that agent, checks it against your policy, and writes a signed record of what it decided.
- A named agent with a human owner. Each install of Claude Desktop on third-party inference gets its own Argent agent identity tied to an accountable person, so no call is anonymous. Identity tethering
- Only the models you allow. The agent can reach only the providers and models your policy names, and a request can be swapped to an approved model. Model control
- Sensitive data handled before it leaves. Prompts and replies are scanned for personal and secret data, then redacted or blocked as your policy says. Sensitive data redaction
- Spend priced and capped. Every request is priced and metered against budgets for the agent, its owner and the teams above them. AI spend caps
- A signed record you can replay. Each decision becomes a hash-chained, signed audit record, and requests are grouped into sessions you can replay. Session replay
How Claude Desktop on third-party inference connects
Claude Desktop on third-party inference reaches Argent in one way.
- Gateway base URL. Claude Desktop on third-party inference points its model URL at your Argent route and uses an Argent agent key in place of the provider key. Argent removes that key and adds the real provider credential from your vault on the way out, so the tool never holds it.
The same steps are in the Argent console under Integrations > AI Clients, and each agent's Connect tab gives the copy-paste values for your tenant.
What you configure
What your team sets in Claude Desktop on third-party inference. These are the settings the Argent console shows for this tool. Values in angle brackets come from your tenant.
- Install the 3P build. Claude Desktop on 3P (separate deployment; local data under Claude-3p). Docs: claude.com/docs/third-party/claude-desktop.
- Open the configuration window. Developer > Configure Third-Party Inference... > Connection: Inference provider = Gateway.
- Gateway fields. Gateway base URL = https://<ARGENT_GATEWAY>/<tenant>/anthropic (the path prefix selects the Argent route). Credential kind = Static API key; Gateway API key = the Argent agent key; Gateway auth scheme = Bearer (default) or x-api-key. Keys: inferenceProvider=gateway, inferenceGatewayBaseUrl, inferenceGatewayApiKey, inferenceGatewayAuthScheme.
- Models. Leave model discovery on, or set inferenceModels to the full model IDs your Argent route allows (an unpriced model is refused).
- Export for the fleet. Click Export to produce a.mobileconfig (macOS) or.reg (Windows) and push it with your MDM (see enterprise section).
- TLS. If the gateway uses a private CA, the app trusts the OS store; on macOS install the CA with full root trust (security add-trusted-cert -d -r trustRoot -k /Library/Keychains/System.keychain ca.pem). Fallback: launchctl setenv NODE_EXTRA_CA_CERTS <bundle> then relaunch.
Roll it out to every machine
Make it the only way Claude Desktop on third-party inference runs on company machines. These controls are documented by the platform each one belongs to (linked); pair them with the Argent settings above.
- MDM profile for Claude Desktop on 3P (Windows). Pins the inference provider to the Argent gateway for every user; users never see a claude.ai login. Vendor docs
- MDM profile for Claude Desktop on 3P (macOS). Same as Windows through a configuration profile. Vendor docs
- Managed settings file (Linux). Same keys for Linux desktops. Vendor docs
- Pin the egress proxy for the app, agent and Cowork shell. Forces the app, the Claude Code engine it runs and (macOS/Windows) the Cowork sandbox shell through one proxy. Vendor docs
- Claude Code managed settings next to the Desktop policy. For the Code tab: proxy/CA variables and merge behaviour. Vendor docs
Argent ships the push kit. For managed fleets the console produces per-MDM scripts (Jamf, Kandji, Intune-style, shell and PowerShell) with your tenant certificate and proxy address built in, plus a fleet setup script and certificate bundle. Your company CA can sign the Argent certificate, so managed machines already trust it.
Want to see which AI tools people already run before you push anything? Start with Shadow AI discovery in watch mode. Watch before you enforce
Check that it works
Send one request and look for the signed row. Send a chat message; a signed LLM ALLOW row naming the agent appears in the Argent console live stream, and cache_read_input_tokens should be above zero on later turns.
Every request also appears on the live stream with the agent's name and owner, and in the session view. Session replay
Where to go next
Keep going. Pages that cover the same ground from another angle.
- Use case: Rolling out copilots and chat safely
- Goals: See all AI in your company, Data loss prevention, Budget and token spend, Model control and routing, Agent identity and ownership
- Capabilities: Keyless agents, Prompt injection defense, Response screening, One-click audit evidence
- Related integrations: LM Studio
- All integrations: Browse every AI client, provider, vault and identity source
Want this running on your own use cases?
Book a demo