Models

Customer aliases and rates.

Use these IDs in POST /v1/chat/completions. The list below comes from the existing GET https://serveras.click/v1/models catalog, refreshed every five minutes; a recent snapshot may appear during a temporary outage. max_output is the customer cap (16,384), not the upstream ceiling.

How to read the rates

The displayed multipliers are credit weights, not RM prices. Credits are internal billing units, not raw model tokens. INPUT weights uncached prompt tokens, CACHE weights eligible cached input, and OUTPUT weights generated tokens.

Every prepaid pack can access the same available models. A larger pack changes how many credits you receive per RM; it does not change model multipliers. See Billing for pack prices, validity, and the complete formula.

curl https://serveras.click/v1/models

Document input

Models whose input_modalities include file accept documents in the same request as text. One limit applies to every model: 100,000 characters of document text per request, with at most 5 documents. A document that exceeds the limit is rejected with 400 before any credit is reserved, so you are never charged for it.

Send the file as a base64 data URL in a file content part. PDF and plain text are both accepted; for PDF the limit applies to the extracted text, so a scanned document costs more tokens than a text one.

curl https://serveras.click/v1/chat/completions \
  -H "Authorization: Bearer $ASAI_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "asai/gpt-6-sol",
    "messages": [{
      "role": "user",
      "content": [
        {"type": "text", "text": "Summarise clause 4."},
        {"type": "file", "file": {
          "filename": "contract.pdf",
          "file_data": "data:application/pdf;base64,JVBERi0xLjQK..."
        }}
      ]
    }]
  }'

Rates are model-specific credit multipliers per token component. Historical settled charges are not recalculated if rates later change.