# Claude Opus 5.5 · API guide

Use the exact API model ID `claude-opus-5-5`. This guide describes the platform contract; check the live catalog before calling the model.

## Request example

The platform origin is filled in at build time. Set API_KEY to your server-side key before running.

```bash
curl "https://vip.astralgames.ai/v1/messages" \
  -H "Content-Type: application/json" \
  -H "x-api-key: $API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  --data '{
  "model": "claude-opus-5-5",
  "max_tokens": 1024,
  "messages": [
    {
      "role": "user",
      "content": "Explain how an HTTP request works in three concise steps."
    }
  ],
  "stream": false
}'
```

## Protocol and response

Platform route limits come from `/v1/models` and may be lower than the official model specifications. The official model uses always-on adaptive thinking and rejects disabled thinking or forced tool selection; review the native parameter rules when upgrading.

This model uses the claude protocol. Read its native response and usage fields; do not wrap the request in the media task input envelope.

Supported endpoint templates：

- `POST /v1/messages`
- `POST /v1/messages/count_tokens`

| Limit | Catalog value |
| --- | --- |
| Context tokens | 128000 |
| Maximum output tokens | 64000 |

For streaming, use the native protocol streaming option and parse its events; check supported capabilities in the live catalog.

<!-- official-reference:start -->
## Official sources and identity

Checked: 2026-09-30 · Platform model: `claude-opus-5-5` · Official reference: `claude-opus-5-5`

- [claude-opus-5-5 official reference](https://platform.claude.com/docs/en/models/opus-5-5/overview)
- [Anthropic Messages API](https://platform.claude.com/docs/en/api/messages/create)
- [Anthropic thinking and sampling rules](https://platform.claude.com/docs/en/build-with-claude/thinking)
<!-- official-reference:end -->

## Pricing and troubleshooting

Check current availability, contract and account pricing before integration. Keep request IDs and usage receipts when diagnosing errors. [Complete API reference](/docs/api)