AnotherWrapper
    $249
    LLMImageVideoSpeechTranscription

    ARC-AGI 2

    🏆 Leaderboard

    As of August 20, 2026, GPT-5.6 Sol is #1 for ARC-AGI 2 at 92.5%. Ranked by ARC-AGI: novel visual puzzles, not memorized exams. 69 models in this index have a published ARC-AGI 2 score. Methodology: ARC-AGI / ARC Prize (https://arcprize.org/). ARC-AGI 2 leaderboard with live API prices. ARC-AGI 2 — second generation abstract reasoning tasks testing general intelligence. Official methodology: ARC-AGI / ARC Prize (https://arcprize.org/).

    Updated August 20, 2026282 models33 providers
    GPT-5.6 Sol
    OpenAI · Proprietary
    		92.5%$35.00
    Claude Opus 5
    Anthropic · Proprietary
    		90.4%$30.00
    Claude Fable 5
    Anthropic · Proprietary
    		89.2%$60.00
    GPT-5.5
    OpenAI · Proprietary
    		85%$35.00
    GPT-5.5 Pro
    OpenAI · Proprietary
    		84.6%$540.00
    GPT-5.6 Terra
    OpenAI · Proprietary
    		83.9%$14.00
    Gemini 3.1 Pro
    Google · Proprietary
    		77.1%$17.50
    Claude Opus 4.7
    Anthropic · Proprietary
    		75.8%$30.00
    GPT-5.4
    OpenAI · Proprietary
    		73.3%$17.50
    Gemini 3.5 Flash
    Google · Proprietary
    		72.1%$10.50
    Claude Opus 4.8
    Anthropic · Proprietary
    		72.1%$30.00
    Claude Opus 4.6
    Anthropic · Proprietary
    		68.8%$30.00
    Grok 4.6
    xAI · Proprietary
    		67.1%$8.00
    DeepSeek-V4-Flash-0731OSS
    DeepSeek · Open Source
    		61.4%$1.76
    Kimi K3OSS
    Moonshot AI · Open Source
    		60.4%$18.00
    Gemini 3.6 Flash
    Google · Proprietary
    		60.4%$4.50
    GPT-5.6 Luna
    OpenAI · Proprietary
    		59.5%$1.40
    Claude Sonnet 4.6
    Anthropic · Proprietary
    		58.3%$18.00
    GPT-5.2 Pro
    OpenAI · Proprietary
    		54.2%$189.00
    GPT-5.2
    OpenAI · Proprietary
    		52.9%$15.75
    Grok 4.5
    xAI · Proprietary
    		52.6%$8.00
    Inkling-SmallOSS
    Thinking Machines · Open Source · via Thinking Machines Lab
    		40.1%$1.50
    Claude Opus 4.5
    Anthropic · Proprietary
    		37.6%$30.00
    Gemini 3 Flash
    Google · Proprietary
    		33.6%$3.50
    Gemini 3 Pro
    Google · Proprietary
    		31.1%$14.00
    GLM-5.2OSS
    Z AI · Open Source
    		22.8%$5.80
    GPT-5.4 mini
    OpenAI · Proprietary
    		18.9%$5.25
    GPT-5.1
    OpenAI · Proprietary
    		17.6%$11.25
    Grok-4
    xAI · Proprietary
    		15.9%$18.00
    Claude Sonnet 4.5
    Anthropic · Proprietary
    		13.6%$18.00
    Kimi K2.5OSS
    Moonshot AI · Open Source
    		11.8%$3.68
    Gemini 3.5 Flash-Lite
    Google · Proprietary
    		10.3%$2.80
    GPT-5 High
    OpenAI · Proprietary
    		9.9%$11.25
    GPT-5
    OpenAI · Proprietary
    		9.9%$11.25
    Claude Opus 4
    Anthropic · Proprietary
    		8.6%$90.00
    GPT-5 Medium
    OpenAI · Proprietary
    		7.5%$11.25
    o3
    OpenAI · Proprietary
    		6.5%$10.00
    o4-mini
    OpenAI · Proprietary
    		6.1%$5.50
    Claude Sonnet 4
    Anthropic · Proprietary
    		5.9%$18.00
    GPT-5.4 nano
    OpenAI · Proprietary
    		5.7%$1.45
    Grok-4 Fast Reasoning
    xAI · Proprietary
    		5.3%$0.70
    Grok 4 Fast
    xAI · Proprietary
    		5.3%$0.70
    Gemini 2.5 Pro
    Google · Proprietary
    		4.9%$11.25
    o3-pro
    OpenAI · Proprietary
    		4.9%$100.00
    MiniMax M2.5OSS
    MiniMax · Open Source
    		4.9%$1.50
    GLM-5OSS
    Z AI · Open Source
    		4.9%$4.20
    GPT-5 mini
    OpenAI · Proprietary
    		4.4%$2.25
    DeepSeek-V3.2OSS
    DeepSeek · Open Source · via OpenRouter
    		4.0%$0.57
    Claude Haiku 4.5
    Anthropic · Proprietary
    		4.0%$6.00
    o3-mini
    OpenAI · Proprietary
    		3.0%$5.50
    Showing 1–50 of 282 models

    Next step

    You found the model. Now ship the product.

    Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.

    Get the stack — $249Have us build it with you

    The index

    All Large Language Models

    282 models across 33 providers. Search or jump to a lab — every model page stays linked here.

    OpenAI51Anthropic22Google25Cursor2xAI15DeepSeek16MiniMax6Qwen38AI21 Labs2Amazon5Baidu1ByteDance2Cohere3IBM1Inception1inclusionAI1LG AI Research1Meituan4Meta12Microsoft5Mistral23Moonshot AI9Nous Research1NVIDIA5Poolside2Sakana AI1StepFun1Tencent1Thinking Machines1Unisound1Upstage1Xiaomi3Z AI20

    Baidu

    1 models
    ERNIE 4.5

    Inception

    1 models
    Mercury 2

    inclusionAI

    1 models
    Ling-3.0-flash

    LG AI Research

    1 models
    K-EXAONE-236B-A23B

    Nous Research

    1 models
    Hermes 3 70B

    Sakana AI

    1 models
    Sakana Namazu

    StepFun

    1 models
    Step-3.5-Flash

    Tencent

    1 models
    Hy3

    Thinking Machines

    1 models
    Inkling-Small

    Unisound

    1 models
    U2

    Upstage

    1 models
    Solar Pro 4

    Which model leads ARC-AGI 2 right now?

    As of August 20, 2026, GPT-5.6 Sol by OpenAI is #1 for ARC-AGI 2 at 92.5%. Ranked by ARC-AGI: novel visual puzzles, not memorized exams. This board also tracks ARC-AGI 2. Next on the same board: Claude Opus 5 and Claude Fable 5. This arc-agi 2 leaderboard ranks models by ARC-AGI 2. Scores come from public evals. Prices are the live API rates in the table above.

    Sources: ARC-AGI / ARC Prize (https://arcprize.org/); OpenAI API pricing (https://developers.openai.com/api/docs/pricing); Anthropic Claude API pricing (https://platform.claude.com/docs/en/about-claude/pricing); Gemini API pricing (https://ai.google.dev/gemini-api/docs/pricing)

    Top 8 for arc-agi 2. Ranked by ARC-AGI: novel visual puzzles, not memorized exams. Input and output are dollars per million tokens.
    RankModelARC-AGI 2Input /MOutput /M
    1GPT-5.6 Sol92.5%$5.00$30.00
    2Claude Opus 590.4%$5.00$25.00
    3Claude Fable 589.2%$10.00$50.00
    4GPT-5.585%$5.00$30.00
    5GPT-5.5 Pro84.6%$60.00$480.00
    6GPT-5.6 Terra83.9%$2.00$12.00
    7Gemini 3.1 Pro77.1%$2.50$15.00
    8Claude Opus 4.775.8%$5.00$25.00

    Related LLM leaderboards

    LLM LeaderboardCodingMathAgenticReasoningWritingScienceFinanceLong ContextVisionOpen Source

    Individual benchmark pages

    Rank one eval at a time. All LLM benchmarks.

    HLESWE BenchHumanity's Last Exam (No Tools)Humanity's Last Exam (With Tools)Terminal-Bench 2.0Terminal-BenchTerminal-Bench 2.1Terminal-Bench 2Terminal-Bench 3.0ARC-AGI-1 Verified

    ARC-AGI 2 FAQ

    Who ranks #1 on the ARC-AGI 2 leaderboard?

    As of August 20, 2026, GPT-5.6 Sol by OpenAI ranks #1 on ARC-AGI 2 at 92.5%. API pricing is $5.00/M input and $30.00/M output.

    What are the top models on ARC-AGI 2?

    The current ARC-AGI 2 ranking as of August 20, 2026 is 1. GPT-5.6 Sol at 92.5%; 2. Claude Opus 5 at 90.4%; 3. Claude Fable 5 at 89.2%.

    Which arc-agi 2 model is the cheapest?

    Llama 4 Scout is the cheapest scored model on this arc-agi 2 leaderboard at $0.08/M input and $0.30/M output ($0.38 blended). GPT-5.6 Sol still leads ARC-AGI 2 at 92.5%.

    Should I always pick the #1 ARC-AGI 2 model?

    Not automatically. GPT-5.6 Sol leads ARC-AGI 2, but a cheaper scored model can be the better production choice if the quality gap is small. Use the table to weigh ARC-AGI 2 against input/output price, context window, and related evals.

    How often is the ARC-AGI 2 leaderboard updated?

    Scores and API prices on this page are refreshed from published evals and provider rates. The snapshot is labeled August 20, 2026. Treat it as a current index, not a one-off blog post.

    What is ARC-AGI 2?

    ARC-AGI 2 — second generation abstract reasoning tasks testing general intelligence. This page ranks models that have published a ARC-AGI 2 score, with live API token prices on the same row. Official methodology: ARC-AGI / ARC Prize (https://arcprize.org/).

    Where is the ARC-AGI 2 leaderboard?

    This page is the ARC-AGI 2 leaderboard. Models are sorted by ARC-AGI 2, with input and output token prices on the same row so you can weigh score against cost. Official boards often omit price; that comparison is the point of this index.

    How is this ARC-AGI 2 ranking different from the official board?

    The official ARC-AGI / ARC Prize page owns the methodology. This page keeps the published ARC-AGI 2 score next to live API $/M so you can pick a production SKU, not only a trophy number. Source: https://arcprize.org/