Quickstart
Three steps to your first request.
Using a coding agent or chat app? Choose your tool and copy its configuration, including model limits and Codex / Claude compatibility.
1. Get a key
Sign in and open API keys in the console. Create a key and copy it. We show it once.
No account yet? Subscribe (€19/month, or €16/month billed yearly at €192/year, for your first year until 1 December; then €29/month or €288/year, excl. VAT) for Chat and API access.
2. Set the key
export VORA_API_KEY=vora_sk_...3. Make a request
curl https://api.voracompute.com/v1/chat/completions \
-H "Authorization: Bearer $VORA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen3.8-27b",
"messages": [
{
"role": "user",
"content": "Say hello in one sentence."
}
],
"reasoning_effort": "medium"
}'Python and Node use the official OpenAI SDK: pip install openai or npm install openai.
Already using the OpenAI SDK?
For Chat Completions, set the base URL, Vora key and model below. Use chat.completions.create(); Vora does not currently expose the Responses API. See agent compatibility if your tool chooses the endpoint for you.
- Base URL:
https://api.voracompute.com/v1 - API key: your
vora_sk_...key - Model:
qwen3.8-27b
Streaming
Set stream to true. Tokens arrive as server-sent events.
curl https://api.voracompute.com/v1/chat/completions -N \
-H "Authorization: Bearer $VORA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen3.8-27b",
"messages": [
{
"role": "user",
"content": "Say hello in one sentence."
}
],
"reasoning_effort": "medium",
"stream": true,
"stream_options": {
"include_usage": true
}
}'The examples include stream_options.include_usage to get token counts. The last event before [DONE] carries usage and has an empty choices array.
{
"model": "qwen3.8-27b",
"messages": [
{
"role": "user",
"content": "Say hello in one sentence."
}
],
"stream": true,
"stream_options": {
"include_usage": true
}
}Optional reasoning
Reasoning defaults to Medium for every chat model. Choose a supported reasoning_effort from the model catalog, or choose Reasoning in Chat. Higher levels allow more thinking and can take longer. Reasoning uses output tokens; it does not guarantee a better answer.
curl https://api.voracompute.com/v1/chat/completions \
-H "Authorization: Bearer $VORA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen3.8-27b",
"messages": [
{
"role": "user",
"content": "Say hello in one sentence."
}
],
"reasoning_effort": "medium"
}'Reasoning is returned in reasoning_content, separately from the final answer in content. The examples print the two separately. See Models & limits for streaming fields, budgets and cancellation.
Next
- Models & limits: what the model supports.
- Errors: what to retry.
- Set up your agent: copy-ready settings, model limits, Codex and Claude compatibility.