Claude Opus 5.5, the newest Anthropic flagship in our catalog, is available on GetVibe API. It went live on September 23, 2026, and since September 24 it accepts the full 1,000,000-token context window. You call it with the same key and endpoint as every other model.
What you get
| Spec | Claude Opus 5.5 |
|---|---|
| Model ID | claude-opus-5.5 |
| Context window | 1,000,000 tokens |
| Max output | 128,000 tokens |
| Input | text and images |
| Reasoning | extended thinking, billed as output tokens |
| Tool calling / JSON mode | yes / yes |
| Price | $0.2027 input · $1.0134 output per 1M tokens |
We tested the full window before publishing it. A request of 732,379 tokens was answered with details from both the start and the end of the prompt, and a request over 1,000,000 tokens was rejected at the model's limit. So the number above is a measured limit, not only the one on the spec sheet.
Call it in one request
The endpoint is OpenAI-compatible. Change the model name and keep the rest of your code:
curl https://getvibeapi.com/v1/chat/completions \
-H "Authorization: Bearer $GETVIBE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-opus-5.5",
"messages": [{"role": "user", "content": "Summarize the risks in this contract: ..."}]
}'If your tooling speaks the Anthropic Messages format (the Anthropic SDK, Claude Code), use /v1/messages on the same host. The setup is in the Claude Code guide and in Messages and Responses.
When the 1M window pays off
- One-shot reading of large inputs: a full codebase dump, a long contract bundle or months of support tickets in a single prompt, without a chunking pipeline.
- Long agent sessions: coding agents collect file contents and tool output fast, and a larger window means fewer forced summaries.
- Cross-document questions: when the answer depends on two passages that are far apart, it helps to have both in context at once.
A full window is expensive per request, because every input token is billed. For repeated prompts that share a long prefix, see prompt caching: cached input reads are billed at a fraction of the input rate. The exact cached price is on the Claude Opus 5.5 page.
Next steps
Compare Claude Opus 5.5 with the rest of the lineup on the Models page, read the reasoning guide to control thinking, or start with the quick-start guide. You pay per token from a prepaid USD balance. There is no subscription.