Claude Sonnet 5.5 API Pricing

Effective AWSQ API rates

USD per 1M tokens. Quota rates use the current AI in IDE base price × 0.8. Provider prices as of 2026-11-02.

Use Claude Sonnet 5.5 in your client

Create a AWSQ API key in your dashboard, replace YOUR_DASHBOARD_API_KEY, and run this request in a Bash-compatible terminal. The key and its funding source need a positive balance; this request consumes quota.

curl --globoff 'https://api.aiinide.com/v1/messages' \
  -H "x-api-key: YOUR_DASHBOARD_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-5-5","messages":[{"role":"user","content":"Hello"}],"max_tokens":256}'

Example: usage versus a top-up purchase

Each row uses a 3:1 input-to-output mix and no cache reads. Quota consumed is the cost of that mix. The top-up column is a separate purchase.

The quota column is what this token mix consumes. The top-up column is the smallest purchase the slider accepts: $50, then steps of $25. A small task does not cost that purchase. Examples round top-up quota up to the next $25 with a $50 minimum. Actual charges vary by usage, plan, context, speed, and promotions. These examples use standard rates and one request, and they do not apply a membership discount. On a billed Claude request the membership discount reduces the token cost first, and the per-request minimum is applied after that discount.

Protocol, channels, and tools

Protocol

Anthropic Messages

Path

/v1/messages

API key header

x-api-key

Request floor

At least USD 0.0004 of quota per billable request on this channel. The membership discount reduces the token cost first; this minimum is applied after that discount, and the channel multiplier after the minimum.

Channels and this model ID

• AWSQ is compatible with this model ID and uses 0.8× the configured model price when the model is in the catalog.
• CCMAX is compatible with this model ID and uses 2× the configured model price when the model is in the catalog.
• CODEX is not compatible with this model ID.
• GEMINI is not compatible with this model ID.

AWSQ and CCMAX also accept OpenAI-compatible clients at the same host with /v1 appended. That is a different path from the Messages example above.

Tool limits on this channel

Web Search max_uses must be 1–20 per request; the default when omitted is 5. One web_fetch tool is accepted, with max_uses from 1 to 20 and max_content_tokens within 1000000. A remote MCP toolset is rejected.

Claude Sonnet 5.5 API FAQ