anthropic
Anthropic: Claude Opus 5.5
anthropic/claude-opus-5.5
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
Anthropic's Claude Opus 5.5, listed on OpenRouter as anthropic/claude-opus-5.5 (canonical anthropic/claude-opus-5.5-20260921), is described by its vendor as a flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It accepts text, image, and file inputs and returns text, with a one-million-token context window. Reasoning is mandatory, with selectable effort levels. OpenRouter's catalog also records leading design-arena placements across several agentic and creative categories, though no independent model card was available in this source pack.[1]
Strengths
- Vendor-reported: Anthropic positions it as its flagship for complex reasoning, coding, and extended agentic work, with particular strength in multi-step changes across large codebases.[1]
- Measured in the catalog: leading design-arena placements in multiple agentic categories (such as full-stack, web apps, and code-to-video) and creative categories (such as game dev and data visualization).[1]
- Broad inputs and tooling per the listing: text, image, and file inputs; tool calling, structured outputs, adjustable reasoning effort, and a one-million-token context window.[1]
Limitations
- Catalog-only evidence: every claim traces to OpenRouter's listing and the vendor description; no independent model card was in the source pack, so validate on your own workloads before adopting.[1]
- Reasoning is mandatory rather than optional; only the effort level is adjustable, which may be unnecessary overhead for simple prompts.[1]
- Evidence gaps: knowledge cutoff is unreported, some evaluation indices are null, and arena coverage skews toward design and agentic tasks rather than broad capability benchmarks.[1]
Where to use it
- Worth evaluating for agentic coding in large repositories, matching the vendor's stated focus on long-horizon, multi-step work.[1]
- Candidate for pipelines combining documents, images, and files with tool calling and structured outputs.[1]
- Plausible fit for design-oriented generation (web and app UI, data visualization) given the arena placements, pending your own testing.[1]
Sources
- [1] OpenRouter catalog · Anthropic: Claude Opus 5.5Sources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
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...
Specifications
- Model ID
- anthropic/claude-opus-5.5
- Canonical ID
- anthropic/claude-opus-5.5-20260921
- Creator
- anthropic
- Added to OpenRouter
- 22.09.2026
- Context length
- 1,000,000
- Modality
- text+image+file->text
- Input modalities
- text, image, file
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $4 / out $20
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | SS — top 10% of the cohort | reasoning: mandatory · default high · efforts: max, xhigh, high, medium, low · parameters: include_reasoning, reasoning, reasoning_effort |
| Coding | No measurement | — |
| Agents | No measurement | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | reasoning: mandatory · default high · efforts: max, xhigh, high, medium, low · parameters: include_reasoning, reasoning, reasoning_effort |
| Document analysis | No measurement | input: file |
| Vision | No measurement | input: image |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Intelligence index | 59.1 | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | agents · androidnativewin rate 76.3% · rank 1 | 1,397Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | agents · blender3dwin rate 57.2% · rank 4 | 1,255Elo | AA — next 20% of the cohort | 13.11.2026 |
| Design Arena | agents · codetovideowin rate 73.8% · rank 1 | 1,341Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | agents · fullstackwin rate 70.8% · rank 1 | 1,392Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | agents · mobileappswin rate 61.6% · rank 3 | 1,296Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | agents · pdfwin rate 68.4% · rank 2 | 1,309Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | agents · spreadsheetwin rate 68% · rank 2 | 1,392Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | agents · webappswin rate 62.7% · rank 1 | 1,320Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · 3dwin rate 65.2% · rank 2 | 1,464Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · asciiartwin rate 64.3% · rank 2 | 1,354Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · codecategorieswin rate 60.8% · rank 2 | 1,385Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · datavizwin rate 63.9% · rank 1 | 1,388Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · gamedevwin rate 63.7% · rank 1 | 1,428Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · svgwin rate 53.6% · rank 3 | 1,321Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · uicomponentwin rate 63.4% · rank 3 | 1,379Elo | SS — top 10% of the cohort | 13.11.2026 |
| Design Arena | models · websitewin rate 55.3% · rank 5 | 1,325Elo | SS — top 10% of the cohort | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| Amazon Bedrock | 1,000,000 | $4 | $20 | — | — | 100% |
| Azure | 1,000,000 | $4 | $20 | — | — | 100% |
| 1,000,000 | $4 | $20 | — | — | 100% | |
| Claude Platform on AWS | 1,000,000 | $4 | $20 | — | — | 100% |
| Anthropic | 1,000,000 | $4 | $20 | — | — | 100% |
| Amazon Bedrock | 1,000,000 | $4.4 | $22 | — | — | 100% |
| Amazon Bedrock | 1,000,000 | $4.4 | $22 | — | — | 100% |
| Azure | 1,000,000 | $4.4 | $22 | — | — | — |
| 1,000,000 | $4.4 | $22 | — | — | — | |
| 1,000,000 | $4.4 | $22 | — | — | 100% | |
| Anthropic | 1,000,000 | $8 | $40 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
- Hark rolls out Handoff², a second-generation browser agent it claims is twice as fastHark began deploying Handoff², the new version of the agent behind its Hark Pro assistant, reporting 82% completion on an internal benchmark, 46% less time per task, and a result within five points of Claude Opus 5.5.
- Show HN: Datatomcp, a data layer where AI agents propose business definitions and people approve themA new Show HN project, datatomcp, pitches itself as an AI-native data layer: databases, sheets and apps are linked once, a person approves the definitions agents use, and every agent question is audited.
- Using AI for fun, plus a roundup of new agent and creative toolsAn RSS excerpt argues that one agent app is enough for expanding what users can do, then collects product updates from Claude, Gemini, Grok Bot, ChatGPT, Granola, Practical Data and Adobe Firefly.
- ExpertCrowd-SWE: Expert-Authored Benchmark Puts Seven Frontier Coding Agents 35 Points ApartA Toptal ExpertCrowd technical report introduces EC-SWE, 100 engineer-written software tasks graded like maintainer review; seven frontier models span 35 points, with Claude Opus 5.5 and GPT-6 Astra statistically tied at the top.
- Google DeepMind and Artificial Analysis Introduce MKA, a Benchmark for Visually Grounded Agent WorkMKA is a 200-task benchmark that measures whether AI agents can extract decisive evidence from blueprints, scans, and charts and turn it into professional deliverables.