AnotherWrapper
    Build with the APIs
    LLMImageVideoSpeechTranscription

    Big Bench Extra Hard

    🏆 Leaderboard

    As of August 31, 2026, Gemma 4 31B is #1 for Big Bench Extra Hard at 74.4%. Ranked by the Big Bench Extra Hard score 11 models in this index have a published Big Bench Extra Hard score. Big Bench Extra Hard leaderboard: rank models by Big Bench Extra Hard next to live API token prices.

    Updated August 31, 2026345 models36 providers

    Big Bench Extra Hard vs price

    11 models. Left is cheaper. Up is a higher score. The line is the best score you can buy at each price — a dot under it is a worse deal than something on the line.

    Best score at each price#1 on this board
    Gemma 4 31BOSS
    Google · Open Source
    		74.4%$0.54
    Gemma 4 26B-A4BOSS
    Google · Open Source
    		64.8%$0.53
    Gemma 4 12BOSS
    Google · Open Source
    		53%$0.53
    DiffusionGemma 26B-A4BOSS
    Google · Open Source
    		47.6%$0.53
    Gemma 4 E4BOSS
    Google · Open Source
    		33.1%$0.06
    Gemma 4 E2BOSS
    Google · Open Source
    		21.9%$0.06
    Gemma 3 27BOSS
    Google · Open Source
    		19.3%$0.30
    Gemma 3 12BOSS
    Google · Open Source
    		16.3%$0.15
    Gemini Diffusion
    Google · Proprietary
    		15%$2.80
    Gemma 3 4BOSS
    Google · Open Source
    		11%$0.06
    Gemma 3 1BOSS
    Google · Open Source
    		7.2%$0.06
    ChatGPT-4o Latest
    OpenAI · Proprietary
    		—$12.50
    Claude 3 Haiku
    Anthropic · Proprietary
    		—$1.50
    Claude 3 Opus
    Anthropic · Proprietary
    		—$90.00
    Claude 3 Sonnet
    Anthropic · Proprietary
    		—$18.00
    Claude 3.5 Haiku
    Anthropic · Proprietary
    		—$4.80
    Claude 3.5 Sonnet
    Anthropic · Proprietary
    		—$18.00
    Claude 3.5 Sonnet
    Anthropic · Proprietary
    		—$18.00
    Claude 3.7 Sonnet
    Anthropic · Proprietary
    		—$18.00
    Claude Fable 5
    Anthropic · Proprietary
    		—$60.00
    Claude Haiku 4.5
    Anthropic · Proprietary
    		—$6.00
    Claude Mythos 5
    Anthropic · Proprietary
    		—$60.00
    Claude Mythos Preview
    Anthropic · Proprietary
    		—$60.00
    Claude Opus 4
    Anthropic · Proprietary
    		—$90.00
    Claude Opus 4.1
    Anthropic · Proprietary
    		—$90.00
    Claude Opus 4.5
    Anthropic · Proprietary
    		—$30.00
    Claude Opus 4.6
    Anthropic · Proprietary
    		—$30.00
    Claude Opus 4.7
    Anthropic · Proprietary
    		—$30.00
    Claude Opus 4.8
    Anthropic · Proprietary
    		—$30.00
    Claude Opus 5
    Anthropic · Proprietary
    		—$30.00
    Claude Sonnet 4
    Anthropic · Proprietary
    		—$18.00
    Claude Sonnet 4.5
    Anthropic · Proprietary
    		—$18.00
    Claude Sonnet 4.6
    Anthropic · Proprietary
    		—$18.00
    Claude Sonnet 5
    Anthropic · Proprietary
    		—$12.00
    Codestral-22BOSS
    Mistral · Open Source · via Mistral AI
    		—$1.20
    Command A+OSS
    Cohere · Open Source
    		—$12.50
    Command R+OSS
    Cohere · Open Source
    		—$1.25
    Composer 2
    Cursor · Proprietary
    		—$3.00
    Composer 2 Fast
    Cursor · Proprietary
    		—$9.00
    DeepSeek R1 Distill Llama 70BOSS
    DeepSeek · Open Source
    		—$0.50
    DeepSeek R1 Distill Qwen 32BOSS
    DeepSeek · Open Source
    		—$0.30
    DeepSeek VL2OSS
    DeepSeek · Open Source · via Replicate
    		—$2.80
    DeepSeek VL2 SmallOSS
    DeepSeek · Open Source · via Replicate
    		—$1.70
    DeepSeek VL2 TinyOSS
    DeepSeek · Open Source · via Replicate
    		—$0.90
    DeepSeek-R1OSS
    DeepSeek · Open Source
    		—$2.74
    DeepSeek-R1-0528OSS
    DeepSeek · Open Source
    		—$2.74
    DeepSeek-V2.5OSS
    DeepSeek · Open Source
    		—$0.42
    DeepSeek-V3OSS
    DeepSeek · Open Source
    		—$1.37
    DeepSeek-V3 0324OSS
    DeepSeek · Open Source
    		—$1.42
    DeepSeek-V3.1OSS
    DeepSeek · Open Source
    		—$1.27
    Showing 1–50 of 345 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.

    Build with the APIsHave us build it with you

    The index

    All Large Language Models

    345 models across 36 providers. Search or jump to a lab — every model page stays linked here.

    OpenAI52Anthropic22Google34Cursor2xAI17DeepSeek20MiniMax7Qwen55AI21 Labs2Amazon7Baidu2ByteDance4Cohere4IBM2Inception1inclusionAI1LG AI Research1Liquid AI2Meituan4Meta12Microsoft9Mistral26Moonshot AI11Nous Research1NVIDIA9OpenBMB1Poolside2Sakana AI1Sarvam AI2StepFun2Tencent1Thinking Machines1Unisound1Upstage1Xiaomi5Z AI21

    Baidu

    2 models
    ERNIE 5.0ERNIE 4.5

    Inception

    1 models
    Mercury 2

    inclusionAI

    1 models
    Ling-3.0-flash

    LG AI Research

    1 models
    K-EXAONE-236B-A23B

    Liquid AI

    2 models
    LFM2.5-VL-3BLFM2.5-2.6B

    Nous Research

    1 models
    Hermes 3 70B

    OpenBMB

    1 models
    MiniCPM-SALA

    Sakana AI

    1 models
    Sakana Namazu

    Sarvam AI

    2 models
    Sarvam-105BSarvam-30B

    Tencent

    1 models
    Hy3

    Thinking Machines

    1 models
    Inkling-Small

    Unisound

    1 models
    U2

    Upstage

    1 models
    Solar Pro 4

    Which model leads Big Bench Extra Hard right now?

    As of August 31, 2026, Gemma 4 31B by Google is #1 for Big Bench Extra Hard at 74.4%. Ranked by the Big Bench Extra Hard score This board also tracks Big Bench Extra Hard. Next on the same board: Gemma 4 26B-A4B and Gemma 4 12B. This big bench extra hard leaderboard ranks models by Big Bench Extra Hard. Scores come from public evals. Prices are the live API rates in the table above.

    Sources: 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 big bench extra hard. Ranked by the Big Bench Extra Hard score Input and output are dollars per million tokens.
    RankModelBig Bench Extra HardInput /MOutput /M
    1Gemma 4 31B74.4%$0.14$0.40
    2Gemma 4 26B-A4B64.8%$0.13$0.40
    3Gemma 4 12B53%$0.13$0.40
    4DiffusionGemma 26B-A4B47.6%$0.13$0.40
    5Gemma 4 E4B33.1%$0.02$0.04
    6Gemma 4 E2B21.9%$0.02$0.04
    7Gemma 3 27B19.3%$0.10$0.20
    8Gemma 3 12B16.3%$0.05$0.10

    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

    Big Bench Extra Hard FAQ

    Who ranks #1 on the Big Bench Extra Hard leaderboard?

    As of August 31, 2026, Gemma 4 31B by Google ranks #1 on Big Bench Extra Hard at 74.4%. API pricing is $0.14/M input and $0.40/M output.

    What are the top models on Big Bench Extra Hard?

    The current Big Bench Extra Hard ranking as of August 31, 2026 is 1. Gemma 4 31B at 74.4%; 2. Gemma 4 26B-A4B at 64.8%; 3. Gemma 4 12B at 53%.

    Which big bench extra hard model is the cheapest?

    Gemma 4 E4B is the cheapest scored model on this big bench extra hard leaderboard at $0.02/M input and $0.04/M output ($0.06 blended). Gemma 4 31B still leads Big Bench Extra Hard at 74.4%.

    Should I always pick the #1 Big Bench Extra Hard model?

    Not automatically. Gemma 4 31B leads Big Bench Extra Hard, but a cheaper scored model can be the better production choice if the quality gap is small. Use the table to weigh Big Bench Extra Hard against input/output price, context window, and related evals.

    How often is the Big Bench Extra Hard leaderboard updated?

    Scores and API prices on this page are refreshed from published evals and provider rates. The snapshot is labeled August 31, 2026. Treat it as a current index, not a one-off blog post.

    What is Big Bench Extra Hard?

    Big Bench Extra Hard is a public LLM eval (the Big Bench Extra Hard score). This page ranks models that have published a score, next to live API prices.

    Where is the Big Bench Extra Hard leaderboard?

    This page is the Big Bench Extra Hard leaderboard. Models are sorted by Big Bench Extra Hard, 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 Big Bench Extra Hard ranking different from the official board?

    Official eval pages own the methodology. This page keeps the published Big Bench Extra Hard score next to live API $/M so you can pick a production SKU, not only a trophy number.