← All models

Claude Opus 5.5 (Max, Default Fallback)

★ featured
Anthropic · released 2026-09-22 · 13 offers

Output 96 tokens/sFirst token 536 sContext 1M tokens

Top 4 cheapest providers (Adjusted $/task)

The same list price can give a different adjusted $/task (caching, token efficiency) — click a price for its inputs.

Within the active global provider, residency and confidentiality filters.

#ProviderAdjusted $/task
1Amazon Bedrockvia OpenRouter — Price via OpenRouter: OpenRouter's price for Amazon Bedrock's endpoint (https://openrouter.ai/anthropic/claude-opus-5.5/providers), collected 2026-11-17.
2Azurevia OpenRouter — Price via OpenRouter: OpenRouter's price for Azure's endpoint (https://openrouter.ai/anthropic/claude-opus-5.5/providers), collected 2026-11-17.
3Googlevia OpenRouter — Price via OpenRouter: OpenRouter's price for Google's endpoint (https://openrouter.ai/anthropic/claude-opus-5.5/providers), collected 2026-11-17.
4T-Systems LLM Hubdirect — Direct price: T-Systems LLM Hub's own price list (https://docs.llmhub.t-systems.net/models/llms/), collected 2026-11-17.

Composite

5 of 7 inputs · 4 from the model family6 radar axes: DesignArena's two boards share one

95.9

includes −0.1 for its Benchmaxxing signal (from 96.0; why, switch off in Options)

AA Coding v4.xCoding Agent v1.4AA IntelligenceAA AgenticEpoch ECISoftware ECIDesignArenaAA Intelligence: percentile 100Epoch ECI: percentile 96Software ECI: percentile 96DesignArena: percentile 96

AA Coding v4.x —Coding Agent v1.4 —AA Intelligence 59.1AA Agentic —Epoch ECI 168.5Software ECI 169.6DesignArena 1319/1389

Radar: percentile among all models measured on each input; a gap means not measured.

Benchmark sheet

75 of 320 registered benchmark versions · bars show the percentile among all models measured on each benchmark.

20 of 76 values are Anthropic's own claims (†), not independent measurements; a matching independent result replaces a claim as soon as one exists.

Composite attachments (used in the score, not counted as exact benchmarks):
  • Epoch ECI · attached
  • Software ECI · attached
  • DesignArena Web Apps (agentic) · attached
  • DesignArena Full-Stack · attached
Compare this model →

Agentic

  • Harvey LAB-AA7791.2%

    Harvey LAB-AA published 2026-09-10 → · Published board — Tests legal-work deliverables across practice areas on Harvey's private task set. Artificial Analysis' run, graded by one LLM judge against task rubrics — not the same run or scale as Vals AI's HLAB row.

  • Terminal-Bench v4.0 (AA, upstream timeouts)9559.6%

    Terminal-Bench v4.0 (AA, upstream timeouts) 4.0-upstream-timeouts → · Published board — Tests terminal-based work on the 66-task release using AA's mini-swe-agent harness with the tasks' own timeouts.

  • EBR-bench (Earthborne Rangers, Epoch AI)9371.4%

    EBR-bench (Earthborne Rangers, Epoch AI) published 2026-09-18 → · Published board — Learning-capability test: models repeatedly play the obscure campaign board game Earthborne Rangers with note-taking, and the score measures whether results improve across playthroughs.

  • Finance Agent v2 (Vals Index v2)8158.6%

    Finance Agent v2 (Vals Index v2) v2 → · Published board — Multi-step financial reasoning tasks.

  • HLAB — Harvey's Legal Agent Benchmark (Vals Index v2)333.8%

    HLAB — Harvey's Legal Agent Benchmark (Vals Index v2) v2 → · Published board — Long-horizon legal work product creation. The index column is Harvey's Legal Agent Benchmark's own published standalone score, reported as accuracy — not comparable with Artificial Analysis' Harvey LAB-AA row.

  • Vals Index v2 (Vals AI)10069.7%

    Vals Index v2 (Vals AI) v2 → · Published board — GDP-weighted average of agentic model performance across finance, coding and legal tasks.

  • Terminal-Bench 4.0developer's claim64.8%† self-reported by the developer

    Terminal-Bench 4.0 v4.0 → · Vendor-reported by Anthropic — Terminal-Bench 4.0 result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic's own run, not an independent measurement.

  • GDPval-AA v2.1developer's claim1,846 Elo† self-reported by the developer

    GDPval-AA v2.1 v2.1 → · Vendor-reported by Anthropic — GDPval-AA v2.1 result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic states the evaluation was run independently by Artificial Analysis; the value is published by Anthropic and kept as the vendor's claim, not as an independent measurement.

  • AutomationBenchdeveloper's claim40.0%† self-reported by the developer

    AutomationBench published 2026-09-22 → · Vendor-reported by Anthropic — AutomationBench result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic states the run was performed and reported by Zapier; the value is published by Anthropic and kept as the vendor's claim, not as an independent measurement.

  • OSWorld 2.0 (partial score)developer's claim81.8%† self-reported by the developer

    OSWorld 2.0 (partial score) v2.0 → · Vendor-reported by Anthropic — OSWorld 2.0 (partial score) result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic's own run, not an independent measurement.

  • OSWorld 2.0 (strict pass rate)developer's claim48.7%† self-reported by the developer

    OSWorld 2.0 (strict pass rate) v2.0 → · Vendor-reported by Anthropic — OSWorld 2.0 (strict pass rate) result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 System Card (PDF, text layer); Anthropic's own run, not an independent measurement.

  • AA-Briefcase v1.1developer's claim1,822 Elo† self-reported by the developer

    AA-Briefcase v1.1 v1.1 → · Vendor-reported by Anthropic — AA-Briefcase v1.1 result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 System Card (PDF, text layer); Anthropic states the evaluation was run independently by Artificial Analysis; the value is published by Anthropic and kept as the vendor's claim, not as an independent measurement.

  • SRE Benchno percentile33.6%

    SRE Bench published 2026-09-22 → · Published board — SRE Bench results published by Vals AI.

  • Terminal-Bench 4.0no percentile61.6%

    Terminal-Bench 4.0 published 2026-09-21 → · Published board — Terminal-Bench 4.0 results published by Vals AI.

  • Terminal-Bench Scienceno percentile48.6%

    Terminal-Bench Science published 2026-09-21 → · Published board — Terminal-Bench Science results published by Vals AI.

  • Terminal-Bench 4.010061.6%

    Terminal-Bench 4.0 published 2026-09-22 → · Published board — Terminal-Bench 4.0 results published by Vals AI.

  • Terminal-Bench Science8848.6%

    Terminal-Bench Science published 2026-09-23 → · Published board — Terminal-Bench Science results published by Vals AI.

  • Tax Agent Bench7870.5%

    Tax Agent Bench published 2026-09-23 → · Published board — Tax Agent Bench results published by Vals AI.

Coding

  • Artificial Analysis Coding Agent Index v2.010061.7%

    Artificial Analysis Coding Agent Index v2.0 v2.0 → · Claude Code — Measures coding-agent systems on FrontierCode v1.1 Main, Terminal-Bench 4.0 and SWE-Atlas-QnA-AA.

  • Artificial Analysis Coding Agent Index v2.110058.2%

    Artificial Analysis Coding Agent Index v2.1 v2.1 → · Claude Code — Measures coding-agent systems on FrontierCode-AA 2.0, Terminal-Bench 4.1 and SWE-Atlas-QnA-AA.

  • SciCode (AA subproblems) v1.0.110066.9%

    SciCode (AA subproblems) v1.0.1 → · Published board — Tests scientific Python programming with scientist-annotated background information.

  • FrontierCode 1.1 Main (Cognition)developer's claim54.4%† self-reported by the developer

    FrontierCode 1.1 Main (Cognition) v1.1 → · Claude Code — Whether a maintainer would merge the agent's pull request, on tasks crafted by open-source maintainers and graded with tests, rubrics and verifiers.

  • Code Migration (Vals Index v2 subset)10068.9%

    Code Migration (Vals Index v2 subset) v2 → · Published board — Porting projects to another language, including COBOL modernization.

  • Vibe Code Bench (Vals Index v2)9790.3%

    Vibe Code Bench (Vals Index v2) v2 → · Published board — End-to-end app-building tasks.

  • FrontierCode v1.1 (Main)developer's claim54.4%† self-reported by the developer

    FrontierCode v1.1 (Main) v1.1 → · Vendor-reported by Anthropic — FrontierCode v1.1 (Main) result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic's own run, not an independent measurement.

  • CursorBench 4.0developer's claim57.8%† self-reported by the developer

    CursorBench 4.0 v4.0 → · Vendor-reported by Anthropic — CursorBench 4.0 result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic's own run, not an independent measurement.

  • SWE-bench Prodeveloper's claim89.9%† self-reported by the developer

    SWE-bench Pro published 2026-09-22 → · Vendor-reported by Anthropic — SWE-bench Pro result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 System Card (PDF, text layer); Anthropic's own run, not an independent measurement.

  • SWE-bench Multilingualdeveloper's claim93.9%† self-reported by the developer

    SWE-bench Multilingual published 2026-09-22 → · Vendor-reported by Anthropic — SWE-bench Multilingual result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 System Card (PDF, text layer); Anthropic's own run, not an independent measurement.

  • SWE-bench Multimodaldeveloper's claim61.4%† self-reported by the developer

    SWE-bench Multimodal published 2026-09-22 → · Vendor-reported by Anthropic — SWE-bench Multimodal result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 System Card (PDF, text layer); Anthropic's own run, not an independent measurement.

  • ProgramBenchno percentile18.5%

    ProgramBench published 2026-09-21 → · Published board — ProgramBench results published by Vals AI.

  • Vibe Code Bench v1.110090.3%

    Vibe Code Bench v1.1 v1.1 → · Published board — Vibe Code Bench v1.1 results published by Vals AI.

  • Code Migration (standalone Vals board, 2026-09-21 snapshot)no percentile55.0%

    Code Migration (standalone Vals board, 2026-09-21 snapshot) published 2026-09-21 → · Published board — Standalone Code Migration results published by Vals AI, distinct from its Vals Index v2 subset.

  • FrontierCode 1.1 Extended (Cognition)developer's claim63.6%† self-reported by the developer

    FrontierCode 1.1 Extended (Cognition) v1.1 → · Claude Code — Whether a maintainer would merge the agent's pull request, on the full 150-task FrontierCode set crafted by open-source maintainers and graded with tests, rubrics and verifiers.

  • ProgramBench10018.5%

    ProgramBench published 2026-09-23 → · Published board — ProgramBench results published by Vals AI.

  • Code Migration (standalone Vals board, 2026-09-22 snapshot)9266.6%

    Code Migration (standalone Vals board, 2026-09-22 snapshot) published 2026-09-22 → · Published board — Standalone Code Migration results published by Vals AI, distinct from its Vals Index v2 subset.

  • LMArena WebDev (Code Arena, overall)1001,827 Elo

    LMArena WebDev (Code Arena, overall) published 2026-09-26 → · Published board — Community pairwise-vote ranking of models on front-end web development tasks, including agentic multi-step coding, published by LMArena as a Bradley-Terry rating.

Efficiency

Knowledge

  • Humanity's Last Exam (AA text-only)10061.4%

    Humanity's Last Exam (AA text-only) published 2026-09-10 → · Published board — Tests expert-level knowledge on AA's text-only Humanity's Last Exam subset.

  • AA-Omniscience Index10046.4

    AA-Omniscience Index published 2026-09-10 → · Published board — Tests factual reliability while rewarding correct answers and penalizing hallucinations.

  • Legal Research Bench (Vals Index v2)9250.5%

    Legal Research Bench (Vals Index v2) v2 → · Published board — Case and statute research with citation-backed answers.

  • Humanity's Last Exam (with tools)developer's claim67.7%† self-reported by the developer

    Humanity's Last Exam (with tools) published 2026-09-22 → · Vendor-reported by Anthropic — Humanity's Last Exam (with tools) result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic's own run, not an independent measurement.

  • Humanity's Last Exam (no tools)developer's claim64.4%† self-reported by the developer

    Humanity's Last Exam (no tools) published 2026-09-22 → · Vendor-reported by Anthropic — Humanity's Last Exam (no tools) result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 System Card (PDF, text layer); Anthropic's own run, not an independent measurement.

  • HealthBench Professionaldeveloper's claim65.6† self-reported by the developer

    HealthBench Professional published 2026-09-22 → · Vendor-reported by Anthropic — HealthBench Professional result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 System Card (PDF, text layer); Anthropic's own run, not an independent measurement.

  • MedScribeno percentile91.4%

    MedScribe published 2026-09-21 → · Published board — MedScribe results published by Vals AI.

  • Public Benefits Bench v1.18670.6%

    Public Benefits Bench v1.1 v1.1 → · Published board — Public Benefits Bench v1.1 results published by Vals AI.

  • MedScribe10091.4%

    MedScribe published 2026-09-22 → · Published board — MedScribe results published by Vals AI.

  • MedCode9049.8%

    MedCode published 2026-09-22 → · Published board — MedCode results published by Vals AI.

Long-context

  • GDP.pdf (AA)no percentile26.2%

    GDP.pdf (AA) published 2026-09-21 → · Published board — Tests professional reasoning over long PDFs with AA document preparation and grading.

  • AA-LCR v1.19984.7%

    AA-LCR v1.1 v1.1 → · Published board — Tests reasoning across multiple long documents with corrected answer keys and grading.

Math

Reasoning

  • ARC-AGI-17497.5%

    ARC-AGI-1 v1 → · Published board — The first-generation ARC-AGI benchmark measuring passive fluid intelligence, with leaderboard score plotted against cost-per-task.

  • ARC-AGI-27891.7%

    ARC-AGI-2 v2 → · Published board — The second-generation ARC-AGI benchmark of passive fluid intelligence, with leaderboard score plotted against cost-per-task.

  • LiveBench 2026-06-259783.2

    LiveBench 2026-06-25 → · Published board — A periodically refreshed suite of objectively graded tasks with separate category scores and an overall category mean.

  • AA Intelligence Index v5.0.1 · snapshot-2026-11-049959.1

    AA Intelligence Index v5.0.1 · snapshot-2026-11-04 → · Published board — Artificial Analysis Intelligence Index v5.0.1. Models AA has not rescored on v5.0.1 are listed as estimates from their frozen v4.x value (isotonic fit over the models measured on both versions), never as measurements.

Science

Tool-use

Vision

  • MMMU Pro (AA)10087.7%

    MMMU Pro (AA) published 2026-09-10 → · Published board — Tests multimodal understanding using challenging ten-option questions.

  • Chartography 100-tasks9666.3%

    Chartography 100-tasks → · Published board — Can a model read the charts professionals rely on — Kaplan-Meier curves, candlesticks, contour maps, Sankey diagrams, Bode plots, wind roses — from the image alone, to an expert-set precision?

  • GDP.pdf 100-tasks8930.6%

    GDP.pdf 100-tasks → · Published board — Can a model answer real professional questions about the PDFs work runs on — manuals, wiring diagrams, filings, forms — so that every rubric criterion an expert wrote is met?

  • Chartography (with tools)developer's claim89.0%† self-reported by the developer

    Chartography (with tools) published 2026-09-22 → · Vendor-reported by Anthropic — Chartography (with tools) result reported by Anthropic for Claude Opus 5.5 in its 2026-09-22 Claude Opus 5.5 launch post; Anthropic's own run, not an independent measurement.

  • Furniture Assembly Benchmark (Epoch AI)10083.3%

    Furniture Assembly Benchmark (Epoch AI) published 2026-09-26 → · Published board — Spatial-reasoning test by Epoch AI: given a photo of a partly or fully assembled IKEA product and its official manual, the model must say whether the build is correct and, if not, name the step with the mistake.

Unusual results

3 threshold-crossing signals flagged · 21 eligible benchmark families

How flags are calculated

Heuristic screen, not statistical significance: benchmark families are correlated and source uncertainty is unknown. Peer evidence needs ≥ 20 independently measured matched configurations from ≥ 10 distinct model families. A flag needs a directed population z-score of magnitude ≥ 1.5 and a gap of ≥ 1.5 from the leave-one-benchmark-family-out mean z in the same direction, over ≥ 5 other benchmark families.

  • unusually weak Vals Index v2 cost per test (Vals AI)
    Why
    Observed: 22.29556 USD
    Peer mean 6.89364 · peer sd 7.83661 · n 38 · families 37
    Directed z -1.965 · baseline z 1.578 · gap -3.543 · profile n 20
    22.29556 USDmeasured· 22.3 USD per rolloutobserved 2026-09-27vals.ai ↗
    Evidence
    Axis: Vals Index v2 cost per test (Vals AI) · v2 · Published board
    Exact value: 22.295559 USD
    Published mean cost: 22.295559 USD per rollout
    Observed: 2026-09-27T05:38:24.953845+00:00 · publication date: not recorded
    Observation id: public:a1e347c697ecbf9cbe1cdd3f
  • unusually strong CritPt (AA) published 2026-09-10
    Why
    Observed: 0.31714 fraction
    Peer mean 0.04412 · peer sd 0.08179 · n 537 · families 370
    Directed z 3.338 · baseline z 1.313 · gap 2.025 · profile n 20
    0.31714 fractionmeasuredobserved 2026-09-22artificialanalysis.ai ↗
    Evidence
    Axis: CritPt (AA) · published 2026-09-10 · Published board
    Exact value: 0.317142857142857 fraction
    Observed: 2026-09-22T22:47:41.526299+00:00 · publication date: not recorded
    Observation id: cr128:88a3953a09bd1380645e0b4a
  • unusually strong Humanity's Last Exam (AA text-only) published 2026-09-10
    Why
    Observed: 0.61353 fraction
    Peer mean 0.15894 · peer sd 0.14559 · n 623 · families 450
    Directed z 3.122 · baseline z 1.324 · gap 1.799 · profile n 20
    0.61353 fractionmeasuredobserved 2026-09-22artificialanalysis.ai ↗
    Evidence
    Axis: Humanity's Last Exam (AA text-only) · published 2026-09-10 · Published board
    Exact value: 0.613531047265987 fraction
    Observed: 2026-09-22T22:47:41.526299+00:00 · publication date: not recorded
    Observation id: cr128:a7ffa5cdde0dd29b43a43391
Missing coverage · 245 benchmark versions

No result does not mean a zero, or that the model was never tested. Collection failures and disputed versions retain their distinct status.

Variants / reasoning settings

Artificial Analysis snapshot 2026-11-04 · Intelligence Index v5.0 (est. = estimated from v4, not measured) · Coding Index v4.x frozen 2026-10-27 · Data: Artificial Analysis · Data: Epoch AI (CC BY)

VariantAA CodingAA Intelligence
Claude Opus 5.5 (Max, Default Fallback)—59.1
Claude Opus 5.5 (Xhigh, Default Fallback)—57.4
Claude Opus 5.5 (High, Default Fallback)—51.8
Claude Opus 5.5 (Medium, Default Fallback)—48.2
Claude Opus 5.5 (Low, Default Fallback)—34.8

SUBSCRIPTION PLAN

GitHub Copilot

Current usage-based billing · model token cost is converted to AI Credits at 1 credit = $0.01.

Input $4.00 / 1M
Cached input $0.200 / 1M
Cache write $5.00 / 1M
Output $20.00 / 1M
Status GA
Token offers by platform · 9 offers (Adjusted $/task)

Click any underlined price to see how it is estimated and where each input comes from. How we calculate adjusted cost.

9 offers within the active global filters, 4 more outside them; “price n/a” means the provider serves the model but publishes no direct token price we could collect — no other channel's price is shown in its place.

OpenRouter (8)

Prices via OpenRouter — what OpenRouter charges for each provider's endpoint, not the provider's own direct price. Source: openrouter.ai/anthropic/claude-opus-5.5/providers, collected 2026-11-17.

Amazon Bedrock
Azure
Google
Amazon Bedrock
Amazon Bedrock
Azure
Google
Google

T-Systems LLM Hub (1)

Direct prices from T-Systems LLM Hub's own price list. Source: docs.llmhub.t-systems.net/models/llms/, collected 2026-11-17.

T-Systems LLM Hub
Claude Opus 5.5 (Max, Default Fallback) — benchmarks & cost