Confirm the model ID
Use current public model pages before larger usage.
Create a project key, use the recommended low-cost text model in the cURL example, send one tiny request against your starter balance, and confirm the result in logs before scaling traffic.
| Step | What to do | Why it matters |
|---|---|---|
| 1 | Create an account and open the console. | You need a project key before sending private API traffic. |
| 2 | Create a project API key and keep it server-side. | Project keys make cost and debugging easier to isolate. |
| 3 | Use the low-cost text model shown below, or copy another current text model from live pricing. | The first call should prove configuration, not stress a volatile free quota pool. |
| 4 | Send one small request, confirm it in logs, then top up only for larger paid-model testing. | This separates integration proof from the purchase decision. |
export TACKLEKEY_API_KEY="replace-with-project-key"
curl https://api.tacklekey.com/v1/chat/completions \
-H "Authorization: Bearer $TACKLEKEY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek/deepseek-v4-flash",
"messages": [{"role": "user", "content": "Reply with one short API health check."}],
"max_tokens": 64
}'
Use current public model pages before larger usage.
Model availability changes; use live pricing before each larger test.
After the starter request works, review the live USD card options only if you need more paid traffic.
Use the 401, 429, invalid model, base URL, and billing checklist.
After the starter request succeeds, review the current USD card options. TackleKey redirects card payments to Creem; the final tax-inclusive amount is shown at Creem checkout. After payment, return to TackleKey to verify the order, credited balance, request log, and one small paid-model request.