RankingGemini 4 Argon

Gemini 4 Argon

Compare models

Data updated 22 Nov 2026

Gemini 4 Argon by Google is a Gemini 4 model, released 2026-09-30. 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.

30 published benchmark measures · 58 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 →

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
Gemini 4 Argon 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 context50100Gemini 4 Argon · Hard reasoning: 78.4 · PreliminaryGemini 4 Argon · Coding: 90.4 · 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 · Preliminary78.4

    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: 93.2 observed win share
      arcprize.org · Source 1
    Coding2 families · 2 with independent evidence · Supported90.4

    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: 82.1–90.6; 2/6 scenarios unsupported. Without one publisher: 88.7–91.7; 2/19 unsupported. Smoothing check: 80.5–94.9; 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: 86.2 observed win share
      tbench.ai · Source 1
    • SWE-bench Pro: 75.0 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
            Google
            Catalog status
            active
            Availability
            Globally available in the paid Gemini API in AI Studio and the Gemini platform, for Google AI Ultra subscribers in Google Antigravity, and for Google AI Ultra and Pro subscribers in the Gemini app and AI Mode.
            Family
            Gemini 4
            Released
            2026-09-30
            Context
            —
            API list price
            $2 input / $10 output per million tokens
            Introductory Gemini API price through 2026-11-30: $2 input and $10 output per million tokens. Cache reads are 95% below the input price. After 2026-11-30, $4 input and $20 output per million tokens.
            License
            proprietary
            Model card
            https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/
            Default Capability family coverage
            Documented-evidence family coverage/badge/gemini-4-argon.svg

            Gemini 4 Argon 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-AnalystAgentHigh48.8%Artificial Analysis
            AA-Briefcase EloHigh1485Artificial Analysis
            AA-Briefcase EloMedium1467Artificial Analysis
            AA-LCR v1.1High79.7%Artificial Analysis
            AA-LCR v1.1Medium78.3%Artificial Analysis
            AA-Omniscience AccuracyHigh48.3%Artificial Analysis
            AA-Omniscience AccuracyMedium49.0%Artificial Analysis
            AA-Omniscience IndexHigh46.1Artificial Analysis
            AA-Omniscience IndexMedium41.1Artificial Analysis
            ARC-AGI-2High91.3%ARC Prize
            ARC-AGI-2Low42.1%ARC Prize
            ARC-AGI-2Medium87.1%ARC Prize
            ARC-AGI-3 · provider-adapterHigh97.3%ARC Prize
            ARC-AGI-3 · provider-adapterLow97.5%ARC Prize
            ARC-AGI-3 · provider-adapterMedium97.3%ARC Prize
            ARC-AGI-3 · standardLow49.8%ARC Prize
            ARC-AGI-3 · standardMedium46.3%ARC Prize
            AutomationBench-AAHigh77.5%Artificial Analysis
            AutomationBench-AAMedium75.8%Artificial Analysis
            CritPtHigh27.1%Artificial Analysis
            CritPtMedium23.1%Artificial Analysis
            DAYJOB FinanceHigh20.3%Surge AI
            DAYJOB HealthcareHigh9.6%Surge AI
            FrontierCode 1.1 · extended-devin-cliHigh59.4%Cognition
            FrontierCode 1.1 · extended-devin-cliLow60.6%Cognition
            FrontierCode 1.1 · extended-devin-cliMedium60.5%Cognition
            FrontierCode 1.1 · main-devin-cliHigh45.9%Cognition
            FrontierCode 1.1 · main-devin-cliLow47.2%Cognition
            FrontierCode 1.1 · main-devin-cliMedium47.5%Cognition
            FrontierCS 2Unspecified54.2%FrontierCS
            FrontierCS 2Unspecified39.6%FrontierCS
            FrontierSWE v2Unspecified55.0%Proximal
            GDP.pdf: All-pass RateHigh21.8%Artificial Analysis
            GDP.pdf: All-pass RateMedium25.6%Artificial Analysis
            GDPval-AA v2.1 LeaderboardHigh1634Artificial Analysis
            GDPval-AA v2.1 LeaderboardMedium1578Artificial Analysis
            Harvey LAB-AA v1.1: Hallucination-Gated All-Pass RateHigh3.9%Artificial Analysis
            Humanity's Last ExamHigh57.1%Artificial Analysis
            Humanity's Last ExamMedium49.9%Artificial Analysis
            LMArena Agent ArenaHigh7.6LMArena
            MLCR-AAHigh13.9%Artificial Analysis
            MMMU-ProMedium88.3%Artificial Analysis
            Multi-step SWEHigh10.1%Klavis AI
            SciCodeHigh61.8%Artificial Analysis
            SciCodeMedium58.9%Artificial Analysis
            sudo L7High18.9%Surge AI
            SWE-Bench Pro V2 Hard · mini-swe-agentHigh92.2%Scale
            Terminal-Bench v4.0High57.1%Artificial Analysis
            Terminal-Bench v4.0Medium55.1%Artificial Analysis
            Terminal-Bench-Science 0.1High52.9%Artificial Analysis
            Terminal-Bench-Science 0.1Medium48.6%Artificial Analysis
            Vals AI · Code Migration v1High68.2%Vals AI
            Vals AI · Excel Modeling Benchmark v1High75.2%Vals AI
            Vals AI · Finance Agent v2High65.4%Vals AI
            Vals AI · Legal Research Bench v1High54.8%Vals AI
            Vals AI · Tax Agent Bench v1: Weighted PassHigh79.0%Vals AI
            Vals AI · Terminal-Bench 2.1High57.7%Vals AI
            Vals AI · Vibe Code Bench v1.1High91.9%Vals AI
            Benchmark scores & sources

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

            BenchmarkBucketScoreHarnessEvidenceSource-recorded date
            SWE-bench ProCoding92.16%SWE-bench Pro reported

            Effort: Not specified

            contributes to capability
            official board2026-10-14
            Terminal-Bench · 4.0Coding57.4%
            Reported settings & source

            Gemini API highest thinking settings; exact named effort not specified by Google. Argon self-computed; other models from the official public leaderboard. Highest scoring thinking level reported by Terminal-Bench authors. Argon agent identity is not specified in this methodology.

            Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed; other models from the official public leaderboard. Highest scoring thinking level reported by Terminal-Bench authors. Argon agent identity is not specified in this methodology. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

            Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 8: Terminal-Bench 4.0 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30

            Published configuration

            Effort: Not specified

            lab self-report

            Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Reviewed 2026-09-30
            Terminal-Bench 4.0Coding54.85%Terminal-Bench 4.0 reported

            Effort: Not specified

            contributes to capability
            official board2026-09-30
            Terminal-Bench-Science · 0.1Coding57.6%
            Reported settings & source

            Gemini API highest thinking settings; exact named effort not specified by Google. Argon self-computed with 6x verifier timeout to address verification timeout issues. Other models from the official public leaderboard. Modified verifier timeout is excluded from matched comparison.

            Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed with 6x verifier timeout to address verification timeout issues. Other models from the official public leaderboard. Modified verifier timeout is excluded from matched comparison. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

            Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 10: Terminal-Bench Science 0.1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30

            Published configuration

            Effort: Not specified

            lab self-report

            Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Reviewed 2026-09-30
            Vals Finance Agent · 2Supporting evidence65.4%
            Reported settings & source

            Gemini API highest thinking settings; exact named effort not specified by Google. Results sourced from Vals AI.

            Provider-published launch claim, not independent measurement by Google for every row. Results sourced from Vals AI. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

            Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 3: Vals Finance Agent 2 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30

            Published configuration

            Effort: Not specified

            lab self-report

            Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Reviewed 2026-09-30
            Vals IndexSupporting evidence68.9%
            Reported settings & source

            Gemini API highest thinking settings; exact named effort not specified by Google. Vals AI composite, weighted by contribution to US GDP; overlapping component benchmarks do not add an aggregate ranking vote.

            Provider-published launch claim, not independent measurement by Google for every row. Vals AI composite, weighted by contribution to US GDP; overlapping component benchmarks do not add an aggregate ranking vote. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

            Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 1: Vals Index / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30

            Published configuration

            Effort: Not specified

            lab self-report

            Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Reviewed 2026-09-30
            Vibe Code Bench · 1.1Supporting evidence91.9%
            Reported settings & source

            Gemini API highest thinking settings; exact named effort not specified by Google. Results sourced from the official Vals AI public leaderboard.

            Provider-published launch claim, not independent measurement by Google for every row. Results sourced from the official Vals AI public leaderboard. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

            Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 7: Vibe Code Bench 1.1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30

            Published configuration

            Effort: Not specified

            lab self-report

            Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

            Reviewed 2026-09-30
            DeepSWE v1.1Agentic————
            GDPval-AAAgentic————
            GPQA DiamondHard reasoning————
            Humanity's Last ExamHard reasoning————
            LiveCodeBenchCoding————
            MMLU-ProKnowledge————
            OSWorld-VerifiedAgentic————
            SWE-bench VerifiedAgentic————
            Terminal-Bench 2.1Agentic————