Claude Sonnet 5.5

What Claude Sonnet 5.5 is

Claude Sonnet 5.5 is a cloud API model for general coding and knowledge work, including structured extraction. It accepts text and image inputs and returns text. It can be used through an application or an eligible Claude surface, but the surrounding product determines the available tools and file workflows.

It is an editorial starting point for a range of ordinary work, not a measured winner in this directory. A model that interprets a quote or drafts a change should still be paired with deterministic checks where possible, such as spreadsheet formulas or a software acceptance test.

How it fits into a workflow

The supplier-comparison workshop illustrates this division of labor: the model interprets quote notes, while formulas calculate totals from preserved inputs. Changing one shipping amount tests whether the workbook really recalculates. The fixture is fictional and the recipe was not run end to end, so it is a useful evaluation design rather than a product success claim.

Examples described as illustrative or proposed trials are editorial suggestions. Documented real builds, when available, are linked separately below.

Requirements, cost and limitations

The documentation lists a 1M-token context and 128K ordinary output limit; tokens are units of text processing, not documents or pages. Migration details include model-specific thinking behavior and unsupported sampling parameters such as temperature. Check current model access and pricing before adapting an older integration.

When to choose it and what to compare

Start with the task rather than an assumed family hierarchy. Haiku may be worth evaluating for short classification or extraction; Opus may be worth evaluating for more complex work. Compare correctness, revisions and total usage on the same examples instead of selecting solely by model size or name.

Access, running cost and practical requirements

Frontier hosted candidate — editorial selection. How we use this label.

Claude API and the documented cloud partner platforms; also a selectable Claude model where the plan allows. Exact plan-by-model entitlement matrix was not independently established.

Developer API prices

Billing route

Claude API base

Units

USD per one million tokens

Input

$2

Cached input

$0.1

Output

$10

5-minute cache write

$2.5

1-hour cache write

$4

Batch input/output discounted 50%; tool/platform charges separate.

API prices are not a chat subscription price. A token is a unit of processed text; images and other inputs have their own token accounting. Repeated history, reasoning output and tools can change the final bill. A subscription may offer separately claimed credits, subject to its terms.

What the configuration changes

Provider positions it for everyday bounded work and faster iteration rather than sustained open-ended judgment.

API default effort high; documented app default medium, so defaults are not interchangeable.

Do not set thinking disabled: use between_tools to suppress upfront thinking. Forced tool use is also rejected.

Provider says Opus remains stronger for complex open-ended work despite some favorable Sonnet benchmark scores.

Benchmark evidence and what it means

These are selected source-reported observations, not a league table. Task, version, tool access, reasoning budget and test harness all affect a result. A historical score does not establish current leadership or accuracy on your own work.

Sources and evidence

Product facts are based on the following published sources, checked 2026-10-27. Workflow suggestions and comparisons are editorial interpretation; PrisAIm has not independently benchmarked this product.