AI Desk · AI FingerprintingNew arrival
Why moderate: near-exact agreement (49 of 49 strings) with anthropic/claude-opus-5, short of a full exact signature.
Anthropic: Claude Opus 5.5 is Anthropic's model. Its measured tokenizer sits with the Claude group, which is evidence that it builds on the Claude tokenizer. Tokenizer reuse is normal engineering, open vocabularies travel between labs, and this says nothing against Anthropic's authorship of the model itself.
Opus 5.5 reads text with the same tokenizer as Opus 5 and Fable 5.1: all 49 measured strings return identical counts, measured on its release day. Its serving envelope matches Fable 5.1 in two further ways. The chat template adds 8 tokens to every request, against 6 on Opus 5, and it held at 8 on all four providers serving it on 22 September 2026 (Anthropic, Amazon Bedrock, Claude Platform on AWS and Azure). Reasoning is mandatory: a request that switches it off is rejected with "Reasoning is mandatory for this endpoint and cannot be disabled", the same answer Fable 5.1 gives, while Opus 5 and Sonnet 5 accept the same request. The price is $4 per million input tokens and $20 per million output, a fifth below Opus 5 for an identical token count, and a cached read costs $0.20 per million, 5% of the input price, against 10% on Opus 5. The base64 row is empty because a content filter stops it before the model runs, with no tokens billed. A bare base64 string, or the same sentence written in hex, is stopped this way on Opus 5.5, Opus 5 and Sonnet 5 alike, and the same text passes once it is wrapped in a request to decode it.
OpenRouter endpoints for Claude Opus 5.5 (snapshot anthropic/claude-opus-5.5-20260921, prices and parameters per provider)OpenRouter model page for Claude Opus 5.5
| We checked | What we found | Where it came from |
|---|---|---|
| How it counts tokens | Counted, using the 49 strings that came back reliably | we measured it |
| How its API is set up | Set up identically to 3 other listings, field for field | api_2aa953b6 |
| How much it can read and write | Reads up to 1,000,000 tokens at once, writes up to 128,000 | its own listing |
| Whether it thinks before answering | Always reasons, and cannot be turned off, at max, xhigh, high, medium, low, high by default | its own listing |
| Who served our requests | Claude Platform on AWS | we measured it, 5 October 2026 |
| Measured twice | The 5 October 2026 reading agreed with the 22 September 2026 reading on every one of the 49 strings clean in both: the tokenizer behind this name held. | 49/49 agree |
Click a row to open the full comparison.
We sent Anthropic: Claude Opus 5.5 fifty short pieces of text and recorded what each one cost it in tokens. The bars below are those costs. Two models built on the same tokenizer produce the same bars; a model built on a different one produces a different set, which is what makes this a fingerprint.
An x marks a row excluded as corrupt (caching or provider interference detected during measurement). Overhead subtracted: 8 prompt tokens.
Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It is particularly strong at multi-step changes in large codebases, code...
Supported parameters
include_reasoningmax_completion_tokensmax_tokensreasoningreasoning_effortresponse_formatstopstructured_outputstemperaturetool_choicetoolsverbosity
defaults: {}
