RankingClaude Sonnet 5.5

Claude Sonnet 5.5

Compare models

Data updated 22 Nov 2026

Claude Sonnet 5.5 by Anthropic is a Claude Sonnet model, released 2026-09-28. This profile collects the published benchmark results and reported effort settings for this model. Configurations are compared on the effort page. Ranks are on the leaderboard.

10 published benchmark measures · 150 measured effort results · 3 benchmark families contribute across 2 task areas. 2 capability estimates available in the full profile. See all results ↓ · Compare published benchmarks →

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare with

Choose a configuration to compare. The overview below combines evidence across settings and has no model-wide rank.

Model evidence summary

This profile combines published settings. It is not a runnable configuration or a leaderboard rank. Compare measured configurations →

Performance profile

Capabilities

Adjusted comparison score · 0–100
Claude Sonnet 5.5 capability radarComparison scores from 0 to 100. Unknown capabilities have no point; lines stop at gaps. Hollow points are preliminary. Exact values and support labels follow the chart.AgenticHard reasoningCodingHuman prefKnowledgeMultimodalLong context50100Claude Sonnet 5.5 · Hard reasoning: 61.0 · PreliminaryClaude Sonnet 5.5 · Coding: 87.5 · Supported

Filled points are supported; hollow points are preliminary. Lines stop at unknown capabilities. Exact values and sources follow below.

AgenticNo comparable evidenceUnknown

0 core families; 0 direct opponents across 0 labs. No admitted core comparisons

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 13/13 scenarios unsupported. Without one publisher: No supported estimate; 15/15 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage; source sensitivity still requires inspection. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

    Hard reasoning1 families · 1 with independent evidence · Preliminary61.0

    1 core families; 59 direct opponents across 10 labs. Needs broader benchmark and opponent support

    Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

    Refit without one family: No supported estimate; 6/6 scenarios unsupported. Without one publisher: No supported estimate; 16/16 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

    Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage. Validation audit.

    Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

    • ARC-AGI: 78.0 observed win share
      arcprize.org · Source 1
    Coding2 families · 2 with independent evidence · Supported87.5

    2 core families; 30 direct opponents across 10 labs.

    Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

    Refit without one family: 77.7–87.8; 2/6 scenarios unsupported. Without one publisher: 84.3–90.6; 2/19 unsupported. Smoothing check: 77.6–92.5; 0/3 unsupported. These are sensitivity checks, not confidence intervals.

    Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage. Validation audit.

    Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

    • Terminal-Bench: 100.0 observed win share
      tbench.ai · Source 1
    • SWE-bench Pro: 37.5 observed win share
      labs.scale.com · Source 1
    Human prefNo comparable evidenceUnknown

    0 core families; 0 direct opponents across 0 labs. No admitted core comparisons

    Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

    Refit without one family: No supported estimate; 1/1 scenarios unsupported. Without one publisher: No supported estimate; 1/1 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

    Audit: cross-family and cross-publisher predictions could not be evaluated from the available comparison network. Validation audit.

    Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

      KnowledgeNo comparable evidenceUnknown

      0 core families; 0 direct opponents across 0 labs. No admitted core comparisons

      Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

      Refit without one family: No supported estimate; 4/4 scenarios unsupported. Without one publisher: No supported estimate; 5/5 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

      Audit: almost no publisher-held-out comparisons remained estimable. Cross-source validity is unresolved. Validation audit.

      Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

        MultimodalNo comparable evidenceUnknown

        0 core families; 0 direct opponents across 0 labs. No admitted core comparisons

        Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

        Refit without one family: No supported estimate; 8/8 scenarios unsupported. Without one publisher: No supported estimate; 8/8 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

        Audit: limited publisher-held-out coverage. Cross-source validity is unresolved. Validation audit.

        Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

          Long contextNo comparable evidenceUnknown

          0 core families; 0 direct opponents across 0 labs. No admitted core comparisons

          Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

          Refit without one family: No supported estimate; 5/5 scenarios unsupported. Without one publisher: No supported estimate; 5/5 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

          Audit: family and publisher holdouts underperformed neutral predictions. Treat this as an exploratory estimate. Validation audit.

          Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

            Scores estimate outcomes against a shared reference panel; they are not accuracy percentages. Sparse or disconnected evidence cannot qualify an overall profile. Open a capability to inspect its evidence.

            Reported effort · Not specified

            Settings reported in this model's published benchmark results, including results outside the aggregate. Effort names are provider-specific. These are not API defaults or equal compute budgets.

            Mixed settings means multiple settings occur in the evidence. Best across efforts means the source selected its best reported result across settings; it does not mean Max. Unspecified settings stay unknown. This model-summary chart combines reported settings. The leaderboard keeps identified configurations separate and excludes unknown effort.

            Inspect each result and its source ↓ · Download effort evidence
            Score contributions and missing evidence

            3 contributing families across 2 capabilities. Fixed reference panels do not change when the catalog expands.

            Capability is fitted jointly across families. Capability estimates below describe different task areas; their weighted sum is not the Capability score.

            Results without reviewed compatibility or a reference match remain in the raw evidence below. Coverage counts only contributing results.

            Model information & shareable badge
            Lab
            Anthropic
            Catalog status
            active
            Availability
            Public provider catalog; account and region restrictions may apply
            Family
            Claude Sonnet
            Released
            2026-09-28
            Context
            1,000,000 tokens
            API list price
            $2 input / $10 output per million tokens
            License
            proprietary
            Model card
            https://platform.claude.com/docs/en/models/sonnet-5-5/overview
            Default Capability family coverage
            Documented-evidence family coverage/badge/claude-sonnet-5-5.svg

            Claude Sonnet 5.5 measured results by effort setting

            Each row is one published result for one effort setting. Evidence collected 2026-11-22. Compare the effort settings →

            BenchmarkEffortScoreSource
            AA-AnalystAgentMax57.5%Artificial Analysis
            AA-Briefcase EloHigh1634Artificial Analysis
            AA-Briefcase EloLow1268Artificial Analysis
            AA-Briefcase EloMax1815Artificial Analysis
            AA-Briefcase EloMedium1438Artificial Analysis
            AA-Briefcase EloXHigh1739Artificial Analysis
            AA-LCR v1.1High78.0%Artificial Analysis
            AA-LCR v1.1Low76.0%Artificial Analysis
            AA-LCR v1.1Max82.7%Artificial Analysis
            AA-LCR v1.1Medium76.3%Artificial Analysis
            AA-LCR v1.1XHigh79.7%Artificial Analysis
            AA-Omniscience AccuracyHigh50.3%Artificial Analysis
            AA-Omniscience AccuracyLow46.0%Artificial Analysis
            AA-Omniscience AccuracyMax51.5%Artificial Analysis
            AA-Omniscience AccuracyMedium45.8%Artificial Analysis
            AA-Omniscience AccuracyXHigh50.9%Artificial Analysis
            AA-Omniscience IndexHigh18.7Artificial Analysis
            AA-Omniscience IndexLow20.3Artificial Analysis
            AA-Omniscience IndexMax30.5Artificial Analysis
            AA-Omniscience IndexMedium18.6Artificial Analysis
            AA-Omniscience IndexXHigh20.6Artificial Analysis
            ARC-AGI-2High67.4%ARC Prize
            ARC-AGI-2Low24.0%ARC Prize
            ARC-AGI-2Medium32.8%ARC Prize
            ARC-AGI-2XHigh83.9%ARC Prize
            ARC-AGI-3 · provider-adapterHigh35.5%ARC Prize
            ARC-AGI-3 · provider-adapterLow0.6%ARC Prize
            ARC-AGI-3 · provider-adapterMedium4.9%ARC Prize
            ARC-AGI-3 · provider-adapterXHigh72.4%ARC Prize
            ARC-AGI-3 · standardHigh1.2%ARC Prize
            ARC-AGI-3 · standardLow0.4%ARC Prize
            ARC-AGI-3 · standardMedium0.4%ARC Prize
            ARC-AGI-3 · standardXHigh24.3%ARC Prize
            AutomationBench-AAHigh59.4%Artificial Analysis
            AutomationBench-AALow49.4%Artificial Analysis
            AutomationBench-AAMax71.8%Artificial Analysis
            AutomationBench-AAMedium54.9%Artificial Analysis
            AutomationBench-AAXHigh65.5%Artificial Analysis
            CritPtHigh24.6%Artificial Analysis
            CritPtLow11.4%Artificial Analysis
            CritPtMax31.4%Artificial Analysis
            CritPtMedium16.9%Artificial Analysis
            CritPtXHigh31.1%Artificial Analysis
            FrontierCode 1.1 · extended-claude-codeHigh61.5%Cognition
            FrontierCode 1.1 · extended-claude-codeLow43.4%Cognition
            FrontierCode 1.1 · extended-claude-codeMax59.1%Cognition
            FrontierCode 1.1 · extended-claude-codeMedium50.5%Cognition
            FrontierCode 1.1 · extended-claude-codeXHigh64.4%Cognition
            FrontierCode 1.1 · main-claude-codeHigh49.4%Cognition
            FrontierCode 1.1 · main-claude-codeLow29.3%Cognition
            FrontierCode 1.1 · main-claude-codeMax46.2%Cognition
            FrontierCode 1.1 · main-claude-codeMedium36.5%Cognition
            FrontierCode 1.1 · main-claude-codeXHigh52.1%Cognition
            FrontierSWE v2Unspecified61.9%Proximal
            GDP.pdf: All-pass RateHigh25.2%Artificial Analysis
            GDP.pdf: All-pass RateLow16.0%Artificial Analysis
            GDP.pdf: All-pass RateMax25.8%Artificial Analysis
            GDP.pdf: All-pass RateMedium20.2%Artificial Analysis
            GDP.pdf: All-pass RateXHigh24.6%Artificial Analysis
            GDPval-AA v2.1 LeaderboardHigh1551Artificial Analysis
            GDPval-AA v2.1 LeaderboardLow1182Artificial Analysis
            GDPval-AA v2.1 LeaderboardMax1838Artificial Analysis
            GDPval-AA v2.1 LeaderboardMedium1326Artificial Analysis
            GDPval-AA v2.1 LeaderboardXHigh1743Artificial Analysis
            Harvey LAB-AA v1.1: Hallucination-Gated All-Pass RateMax2.8%Artificial Analysis
            HLE · full · Scale official evaluationXHigh46.1%Scale
            HLE · text only · Scale official evaluationXHigh46.6%Scale
            Humanity's Last ExamHigh45.8%Artificial Analysis
            Humanity's Last ExamLow36.2%Artificial Analysis
            Humanity's Last ExamMax55.0%Artificial Analysis
            Humanity's Last ExamMedium39.8%Artificial Analysis
            Humanity's Last ExamXHigh50.0%Artificial Analysis
            LiveBench · AMPS Hard · 2026-06-25Max98.0%LiveBench
            LiveBench · AMPS Hard · 2026-06-25XHigh98.0%LiveBench
            LiveBench · code completion · 2026-06-25Max87.0%LiveBench
            LiveBench · code completion · 2026-06-25XHigh84.8%LiveBench
            LiveBench · code generation · 2026-06-25Max95.8%LiveBench
            LiveBench · code generation · 2026-06-25XHigh93.0%LiveBench
            LiveBench · connections · 2026-06-25Max99.3%LiveBench
            LiveBench · connections · 2026-06-25XHigh98.5%LiveBench
            LiveBench · consecutive events · 2026-06-25Max40.6%LiveBench
            LiveBench · consecutive events · 2026-06-25XHigh90.9%LiveBench
            LiveBench · integrals with game · 2026-06-25Max97.0%LiveBench
            LiveBench · integrals with game · 2026-06-25XHigh100.0%LiveBench
            LiveBench · javascript · 2026-06-25Max77.3%LiveBench
            LiveBench · javascript · 2026-06-25XHigh36.4%LiveBench
            LiveBench · logic with navigation · 2026-06-25Max82.0%LiveBench
            LiveBench · logic with navigation · 2026-06-25XHigh80.0%LiveBench
            LiveBench · math comp · 2026-06-25Max97.1%LiveBench
            LiveBench · math comp · 2026-06-25XHigh97.1%LiveBench
            LiveBench · olympiad · 2026-06-25Max92.5%LiveBench
            LiveBench · olympiad · 2026-06-25XHigh91.7%LiveBench
            LiveBench · paraphrase · 2026-06-25Max48.2%LiveBench
            LiveBench · paraphrase · 2026-06-25XHigh62.3%LiveBench
            LiveBench · plot unscrambling · 2026-06-25Max48.6%LiveBench
            LiveBench · plot unscrambling · 2026-06-25XHigh67.8%LiveBench
            LiveBench · python · 2026-06-25Max35.0%LiveBench
            LiveBench · python · 2026-06-25XHigh35.0%LiveBench
            LiveBench · simplify · 2026-06-25Max50.9%LiveBench
            LiveBench · simplify · 2026-06-25XHigh72.4%LiveBench
            LiveBench · spatial · 2026-06-25Max98.0%LiveBench
            LiveBench · spatial · 2026-06-25XHigh98.0%LiveBench
            LiveBench · story generation · 2026-06-25Max69.9%LiveBench
            LiveBench · story generation · 2026-06-25XHigh77.1%LiveBench
            LiveBench · summarize · 2026-06-25Max58.2%LiveBench
            LiveBench · summarize · 2026-06-25XHigh70.3%LiveBench
            LiveBench · tablejoin · 2026-06-25Max39.8%LiveBench
            LiveBench · tablejoin · 2026-06-25XHigh46.9%LiveBench
            LiveBench · tablereformat · 2026-06-25Max98.0%LiveBench
            LiveBench · tablereformat · 2026-06-25XHigh98.0%LiveBench
            LiveBench · theory of mind · 2026-06-25Max86.5%LiveBench
            LiveBench · theory of mind · 2026-06-25XHigh69.2%LiveBench
            LiveBench · typescript · 2026-06-25Max56.7%LiveBench
            LiveBench · typescript · 2026-06-25XHigh46.7%LiveBench
            LiveBench · typos · 2026-06-25Max86.0%LiveBench
            LiveBench · typos · 2026-06-25XHigh84.0%LiveBench
            LiveBench · zebra puzzle · 2026-06-25Max100.0%LiveBench
            LiveBench · zebra puzzle · 2026-06-25XHigh100.0%LiveBench
            LMArena Agent ArenaMax12.5LMArena
            MaxxBenchMax66.9%unLAP
            MLCR-AAHigh18.9%Artificial Analysis
            MLCR-AALow19.4%Artificial Analysis
            MLCR-AAMax41.7%Artificial Analysis
            MLCR-AAMedium17.8%Artificial Analysis
            MLCR-AAXHigh20.6%Artificial Analysis
            SciCodeHigh53.7%Artificial Analysis
            SciCodeLow49.1%Artificial Analysis
            SciCodeMax61.0%Artificial Analysis
            SciCodeMedium52.9%Artificial Analysis
            SciCodeXHigh57.3%Artificial Analysis
            sudo L7Max41.7%Surge AI
            SWE-Bench Pro V2 Hard · mini-swe-agentXHigh70.6%Scale
            Terminal-Bench 4.0 · claudecodeMax61.8%Terminal-Bench
            Terminal-Bench v4.0High43.9%Artificial Analysis
            Terminal-Bench v4.0Low20.7%Artificial Analysis
            Terminal-Bench v4.0Max63.6%Artificial Analysis
            Terminal-Bench v4.0Medium29.8%Artificial Analysis
            Terminal-Bench v4.0XHigh57.1%Artificial Analysis
            Terminal-Bench-Science 0.1High34.3%Artificial Analysis
            Terminal-Bench-Science 0.1Low10.5%Artificial Analysis
            Terminal-Bench-Science 0.1Max53.3%Artificial Analysis
            Terminal-Bench-Science 0.1Medium16.2%Artificial Analysis
            Terminal-Bench-Science 0.1XHigh52.4%Artificial Analysis
            Vals AI · Code Migration v1Max69.8%Vals AI
            Vals AI · Excel Modeling Benchmark v1Max75.7%Vals AI
            Vals AI · Finance Agent v2Max58.1%Vals AI
            Vals AI · Legal Research Bench v1Max48.1%Vals AI
            Vals AI · Tax Agent Bench v1: Weighted PassMax73.4%Vals AI
            Vals AI · Terminal-Bench 2.1High83.1%Vals AI
            Vals AI · Vibe Code Bench v1.1Max92.4%Vals AI
            Benchmark scores & sources

            Original results, evaluation harnesses, and evidence behind this model.

            BenchmarkBucketScoreHarnessEvidenceSource-recorded date
            ARC-AGI-2Hard reasoning83.9%ARC Prize verified

            Effort: Not specified

            contributes to capability
            official board2026-09-28
            AutomationBench-AASupporting evidence71.76882128193995%Artificial Analysis AutomationBench-AA

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-11-21
            Bug Hunt BenchSupporting evidence48.86%Bug Hunt Bench · max

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-09-29
            Bug Hunt BenchSupporting evidence34.29%Bug Hunt Bench · xhigh

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-09-29
            DAYJOB FinanceSupporting evidence40.5%DAYJOB Finance · max

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-11-20
            DAYJOB HealthcareSupporting evidence34%DAYJOB Healthcare · max

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-11-20
            FrontierCS 2Supporting evidence50%Claude Code

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-11-20
            FrontierSWE v2Supporting evidence61.9%Proximal FrontierSWE v2 · Proximus · unspecified

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-11-22
            LMArena Agent ArenaSupporting evidence15.56 LMArena Agent Arena · max

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-11-18
            LMArena Agent ArenaSupporting evidence7.42 LMArena Agent Arena · high

            Effort: Not specified

            official board

            Supporting evidence outside the reviewed capability core

            2026-11-18
            SWE-bench ProCoding70.59%SWE-bench Pro reported

            Effort: Not specified

            contributes to capability
            official board2026-10-12
            Terminal-Bench 4.0Coding61.82%Terminal-Bench 4.0 reported

            Effort: Not specified

            contributes to capability
            official board2026-09-28
            DeepSWE v1.1Agentic————
            GDPval-AAAgentic————
            GPQA DiamondHard reasoning————
            Humanity's Last ExamHard reasoning————
            LiveCodeBenchCoding————
            LMArena Text ArenaHuman pref————
            MMLU-ProKnowledge————
            OSWorld-VerifiedAgentic————
            SWE-bench VerifiedAgentic————
            Terminal-Bench 2.1Agentic————