Set up SpiderGate with your own keys
Who this is for: Anyone who just signed up for SpiderGate, or whose requests come back with no_access, and wants to know which provider keys to add and how.
What this does: Walks you through the six-step setup that picks your keys, adds them, asks a teammate for the ones they hold, and shows you how to plug in your tools. When you finish, the Models page shows how many models you can use.
Before you begin
You can sign in at
app.spideriq.ai. No account yet? Create your account first. Setup belongs to the workspace you are signed in as, so pick the right one in the workspace switcher first.SpiderGate runs on keys you bring. A workspace with no key of its own can browse every model but cannot call one: the gateway answers
403 no_accessand names the keys that would unlock the model. See Errors.To share free keys with the community pool or change that choice, you need the admin role on the workspace. Anyone in the workspace can add keys and walk the setup.
Steps
1. Open setup
Go to app.spideriq.ai/dashboard/gate/setup. You also get there from the setup bubble in the bottom-right corner of every dashboard page, or from Continue setup on the screen you see after you sign in while setup is unfinished. The top of the page shows the six steps: How you'll use it, What you pay for, Free keys, Teammates, Add keys, Ready.
2. Say how you will use SpiderGate
Tick every card that fits: Connect my AI agents, Chat and test models, Build my own app, Make images, video and voice, Share AI access with my team. If you tick agents, pick the tools you use (OpenClaw, Hermes, Cursor, n8n or your own code). Click Continue. Your answers decide which keys the next steps suggest and which connection steps you see at the end.
3. Tick what you already pay for
The page lists subscriptions first, then API keys. Tick the ones you already have: those keys go to the top of the queue on the Add keys step. Leave it blank if you pay for nothing yet; free keys come next.
4. Pick the free keys you want
Each card is a free key, with its free allowance and whether the provider asks for a phone check or a card on file. Untick any you do not want. The Share your free keys with the community pool box is ticked for new free keys: a free key you share serves other workspaces from its own free quota, and while you share at least one active free key your workspace can use the pool too. Untick it to keep your free keys private. Paid and subscription keys always stay private.
5. Ask a teammate for keys they hold
If someone on your team already has a key, add their email and tick the providers to ask them for. Click Send invites & continue. Each person gets an email with a link to a page that shows a guide for each provider, checks the key live before saving it, and never shows the key back. Their keys land in your vault, and they do not need an account. Click Skip if you have nobody to ask.
6. Add your keys
The queue on the left groups what you picked into Subscriptions, API keys and Free keys, with a count of each. For each one, open the provider's page from the guide, create the key, paste it and save. How we check it says what SpiderGate tests before it stores the key. Skip moves to the next one. The full guide for each provider is also under Provider guides.
7. Connect your tools
The last step reads You can now use N models and lists what is connected and who you are still waiting on. Under Connect your agents, copy the Base URL https://spideriq.ai/api/gate/v1 and follow the line for your tool. Any OpenAI-compatible tool works the same way.
8. Leave and come back any time
Save & exit on every step keeps your place. The setup bubble shows the step you stopped on; you can minimise it or choose not to show it again. After you sign in, an unfinished setup offers Continue setup or Go to dashboard, with a box to stop asking at sign-in.
Verify it worked
Open app.spideriq.ai/dashboard/gate/models. The banner reads "You can use N models today." with N above 0. Send one request with a model from that list:
curl https://spideriq.ai/api/gate/v1/chat/completions \
-H "Authorization: Bearer $SPIDERIQ_TOKEN" \
-H "Content-Type: application/json" \
-d '{"model": "spideriq/chat", "messages": [{"role": "user", "content": "Say hello"}]}'A 200 with a reply means your keys serve the request. A 403 whose error.code is no_access means none of your keys can serve that model yet: the error names the providers that would.
Troubleshoot
The Models page still says I can use 0 models? No key was saved, or the one you saved failed its check. Open the Vault and look at the key's status; add it again from the provider's guide if the check failed.
I get
no_accessfor a model I can see? Your keys do not reach that model's provider. Add a key for one of the providers the error names, or pick a model from the list you can use.I get
503 capability_unavailable? SpiderGate could not check which models your workspace may use. Wait a moment and send the request again.The share box is greyed out or missing? Only a workspace admin can change it. If your admin opted the workspace out of the pool, new keys start private.
My teammate's key never arrived? The invite has a time limit shown on their page. Send a new one from the Teammates step, or ask them to add the key from the invite before it runs out.
My agent wants to run setup for me? It can. With the SpiderGate skills installed, an agent reads where setup stands with
gate_onboarding_statusand tells you the next step. It can suggest a teammate to invite, but a person sends the invite.