← All models

Claude Opus 5.5 (Xhigh, Default Fallback)

★ featured
Anthropic · released 2026-09-22 · 8 offers

Output 74 tokens/sFirst token 69 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 2026-11-21.
2Azurevia OpenRouter — Price via OpenRouter: OpenRouter's price for Azure's endpoint (https://openrouter.ai/anthropic/claude-opus-5.5/providers), collected 2026-11-21.
3Googlevia OpenRouter — Price via OpenRouter: OpenRouter's price for Google's endpoint (https://openrouter.ai/anthropic/claude-opus-5.5/providers), collected 2026-11-21.
4T-Systems LLM Hubdirect — Direct price: T-Systems LLM Hub's own price list (https://docs.llmhub.t-systems.net/models/llms/), collected 2026-11-21.

Composite

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

96.6

includes −0.1 for its Benchmaxxing signal (from 96.7; why, switch off in Options)

AA Coding v4.xCoding Agent v1.4AA IntelligenceAA AgenticEpoch ECISoftware ECIDesignArenaAA Intelligence: percentile 100Epoch ECI: percentile 96Software ECI: percentile 99DesignArena: percentile 96

AA Coding v4.x —Coding Agent v1.4 —AA Intelligence 57.4AA Agentic —Epoch ECI 168.7Software ECI 169.8DesignArena 1319/1389

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

Benchmark sheet

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

5 of 26 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-AA7991.2%

    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.

  • Terminal-Bench 4.0developer's claim66.4%† 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.

Coding

  • SciCode (AA subproblems) v1.0.19965.0%

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

  • FrontierCode 1.1 Main (Cognition)developer's claim51.4%† 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.

  • KernelBench-CUDA: DeepSeek NSA (RTX PRO 6000)100109.6

    KernelBench-CUDA: DeepSeek NSA (RTX PRO 6000) rtx-pro-6000 → · Published board — Implement DeepSeek’s Native Sparse Attention block (block top-n routing + sparse attention) as a CUDA kernel over a frozen shape sweep.

  • KernelBench-CUDA: GLM-5.2 Fused MoE (RTX PRO 6000)10010.4

    KernelBench-CUDA: GLM-5.2 Fused MoE (RTX PRO 6000) rtx-pro-6000 → · Published board — Fuse GLM-5.2’s MoE forward pass (256+shared experts, top-8 routing) as one CUDA kernel over a frozen shape sweep; Triton, vLLM and other DSLs are banned — CUDA/PTX/CUTLASS only.

  • KernelBench-CUDA: Grid MinGRU SPS (RTX PRO 6000)8679.4

    KernelBench-CUDA: Grid MinGRU SPS (RTX PRO 6000) rtx-pro-6000 → · Published board — Non-LLM RL simulation: maximise simulation steps per second for a grid world with a MinGRU agent (roofline anchored at 150M peak SPS); fusion optional.

  • KernelBench-CUDA: MegaQwen Decode (RTX PRO 6000)1007.39

    KernelBench-CUDA: MegaQwen Decode (RTX PRO 6000) rtx-pro-6000 → · Published board — Improve the known MegaQwen CUDA megakernel geometry for decode-only token throughput at context lengths 2k–128k (roofline anchored on that decode-only tok/s ceiling); prefill is untimed.

  • Vibe Code Bench 1-10010030.4%

    Vibe Code Bench 1-100 published 2026-09-22 → · Published board — Vibe Code Bench 1-100 results published by Vals AI.

  • FrontierCode 1.1 Extended (Cognition)developer's claim63.5%† 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.

Efficiency

Knowledge

Long-context

  • GDP.pdf (AA)no percentile26.6%

    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.

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-28992.5%

    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 2026-06-258982.1

    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-2026-11-049757.4

    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

  • AutomationBench-AA v1.0.69665.0%

    AutomationBench-AA v1.0.6 → · Published board — Tests multi-app SaaS workflows through REST tools on a held-out AutomationBench split.

Vision

Unusual results

1 threshold-crossing signal flagged · 10 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 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 1.614 · gap 1.725 · profile n 9
    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: cr128:6bd2a51a630d3fdd0ef4cb72
Missing coverage · 295 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 · 4 offers (Adjusted $/task)

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

4 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 (3)

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 2026-11-21.

Amazon Bedrock
Azure
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 2026-11-21.

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