← All models

Grok 4.6 (Xhigh)

deprecated by benchmark source
xAI · released 2026-08-12 · 4 offers

Context 500K tokens

Top 2 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
1AWS Bedrockdirect — Direct price: AWS Bedrock's own price list (https://aws.amazon.com/bedrock/pricing/), collected 2026-11-21.
2Google Vertex AIdirect — Direct price: Google Vertex AI's own price list (https://cloud.google.com/vertex-ai/generative-ai/pricing), collected 2026-11-21.

Composite

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

64.0

includes −5.8 for its Benchmaxxing signal (from 69.8; why, switch off in Options)

Benchmaxxing signal +5.8, light Benchmaxxing tag, uncertain: The 80 % interval reaches below zero — treat this tag as uncertain · light report →The 80 % interval reaches below zero — treat this tag as uncertain.

AA Coding v4.xCoding Agent v1.4AA Intelligence (est.)AA AgenticEpoch ECISoftware ECIDesignArenaAA Coding v4.x: percentile 91AA Intelligence (est.): percentile 93Epoch ECI: percentile 66Software ECI: percentile 56DesignArena: percentile 70

AA Coding v4.x 75.9Coding Agent v1.4 —AA Intelligence (est.) 33.9AA Agentic —Epoch ECI 157.8Software ECI 158.3DesignArena 1231/1237

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

Benchmark sheet

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

2 of 33 values are xAI'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

  • AA-Briefcase941,545 Elo

    AA-Briefcase published 2026-09-10 → · Published board — Tests multi-week professional knowledge-work projects with linked tasks and large source collections.

  • GDPval-AA v2971,663 Elo

    GDPval-AA v2 v2 → · Published board — Tests professional knowledge-work deliverables across occupations using AA's Stirrup harness.

  • Terminal-Bench v2.1 (AA)9588.0%

    Terminal-Bench v2.1 (AA) v2.1 → · Published board — Tests terminal-based work on the 89-task verified refresh using Terminus 2.

  • Terminal-Bench v4.0 (AA)7917.2%

    Terminal-Bench v4.0 (AA) v4.0 → · Published board — Tests terminal-based work on the 66-task release using mini-SWE-agent v2.4.6.

  • EBR-bench (Earthborne Rangers, Epoch AI)4730.5%

    EBR-bench (Earthborne Rangers, Epoch AI) published 2026-09-18 → · Published board — Learning-capability test: models repeatedly play the obscure campaign board game Earthborne Rangers with note-taking, and the score measures whether results improve across playthroughs.

  • RSI-Exam 0.1 (aiming-lab)no percentile0.37

    RSI-Exam 0.1 (aiming-lab) v0.1 → · Grok Build — An agent spends a long, budgeted run improving an inherited, working-but-weak executable research method or harness on a visible split, and the single artifact it submits is re-run once on a sealed hidden split across 88 executable research tasks in six domains.

  • LMArena Agent Arena (overall net improvement)161.2%

    LMArena Agent Arena (overall net improvement) published 2026-09-26 → · Published board — Ranking of models as the orchestrator in LMArena’s Agent Mode, scored by the causal net improvement over the average model across behavioural signals from real sessions.

Coding

  • SciCode (AA subproblems) v1.0.16853.0%

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

  • CursorBench 4.0 (Cursor)developer's claim41.4%† self-reported by the developer

    CursorBench 4.0 (Cursor) v4.0 → · Published board — Agent evaluation on ambiguous, multi-file tasks drawn from real Cursor sessions.

  • DeepSWE (Datacurve, via Epoch AI)6366.7%

    DeepSWE (Datacurve, via Epoch AI) published 2026-09-15 → · mini-SWE-agent — Pass rate of coding agents on original, long-horizon software engineering tasks, run by Datacurve with the mini-swe-agent harness.

  • FrontierSWE v24325.3%

    FrontierSWE v2 v2 → · proximus — 34 hand-written real-world software tasks: every model is evaluated with the site’s own Proximus harness at its maximum reasoning effort, 5 trials per task under a 20-hour budget, scored by the site’s own review protocol; the headline value is mean@5 in percent with best@5/worst@5 bounds.

  • KernelBench-CUDA: GLM-5.2 Fused MoE (RTX PRO 6000)509.39

    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: MegaQwen Decode (RTX PRO 6000)675.42

    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.

  • AA Coding Index v4.x · frozen 2026-10-279175.9

    AA Coding Index v4.x · frozen 2026-10-27 → · Published board — Artificial Analysis withdrew the Coding Index with Intelligence Index v5.0 (2026-10-28). These are the last v4.x values, frozen with their own date (2026-10-27).

Efficiency

  • CursorBench 4.0 cost per task (Cursor)developer's claim$6.10† self-reported by the developer

    CursorBench 4.0 cost per task (Cursor) v4.0 → · Published board — Cursor's published USD cost per task for each CursorBench 4.0 model-effort configuration.

Knowledge

Long-context

  • GDP.pdf (AA)6817.2%

    GDP.pdf (AA) published 2026-09-10 → · Published board — Tests professional reasoning over long PDFs with AA document preparation and grading.

  • AA-LCR v1.19181.0%

    AA-LCR v1.1 v1.1 → · Published board — Tests reasoning across multiple long documents with corrected answer keys and grading.

  • Context Arena — MRCR v2 (8-needle)5581.5%

    Context Arena — MRCR v2 (8-needle) 8-needle → · Published board — The model must find and reproduce the right one of eight near-identical earlier replies hidden in a long conversation, averaged over context lengths from 8k to 128k tokens.

Math

  • FrontierMath Tier 4 v2 (Epoch AI)4331.7%

    FrontierMath Tier 4 v2 (Epoch AI) v2 → · Published board — The hardest, research-level tier of Epoch AI's unpublished FrontierMath problems, run by Epoch AI on its private v2 set.

  • FrontierMath Tiers 1–3 v2 (Epoch AI)5066.0%

    FrontierMath Tiers 1–3 v2 (Epoch AI) v2 → · Published board — Unpublished, expert-written mathematics problems from undergraduate to research level with automatically checkable answers, run by Epoch AI on its private v2 set.

Reasoning

  • Chess Puzzles (Epoch AI)6031.0%

    Chess Puzzles (Epoch AI) published 2026-09-18 → · Published board — Best-move selection on 100 novel chess positions generated programmatically by Epoch AI, each with a single Stockfish-verified best move; probes spatial reasoning and planning.

  • Mystery Game Puzzles (Epoch AI)6434.0%

    Mystery Game Puzzles (Epoch AI) published 2026-09-18 → · Published board — Best-move selection on 100 mid-game positions of a well-known game whose identity Epoch deliberately keeps undisclosed, generated programmatically like Chess Puzzles.

Science

Tool-use

  • AutomationBench-AA v1.0.69867.0%

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

  • τ³-Banking (AA) v1.0.19143.3%

    τ³-Banking (AA) v1.0.1 → · Published board — Tests banking support agents that retrieve policies and change account state through tools.

Vision

  • Chartography 100-tasks2420.5%

    Chartography 100-tasks → · Published board — Can a model read the charts professionals rely on — Kaplan-Meier curves, candlesticks, contour maps, Sankey diagrams, Bode plots, wind roses — from the image alone, to an expert-set precision?

  • GDP.pdf 100-tasks2317.2%

    GDP.pdf 100-tasks → · Published board — Can a model answer real professional questions about the PDFs work runs on — manuals, wiring diagrams, filings, forms — so that every rubric criterion an expert wrote is met?

  • Furniture Assembly Benchmark (Epoch AI)5040.0%

    Furniture Assembly Benchmark (Epoch AI) published 2026-09-26 → · Published board — Spatial-reasoning test by Epoch AI: given a photo of a partly or fully assembled IKEA product and its official manual, the model must say whether the build is correct and, if not, name the step with the mistake.

Missing coverage · 288 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
Grok 4.6 (High)76.833.9est.
Grok 4.6 (Xhigh)75.933.9est.
Grok 4.6 (Medium)74.433.0est.
Grok 4.6 (Low)66.324.2est.

SUBSCRIPTION PLAN

GitHub Copilot

Current usage-based billing · model token cost is converted to AI Credits at 1 credit = $0.01.

Input $2.00 / 1M
Cached input $0.500 / 1M
Output $6.00 / 1M
Status GA
Token offers by platform · 2 offers (Adjusted $/task)

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

2 offers within the active global filters, 2 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.

AWS Bedrock (1)

Direct prices from AWS Bedrock's own price list. Source: aws.amazon.com/bedrock/pricing/, collected 2026-11-21.

AWS Bedrock

Google Vertex AI (1)

Direct prices from Google Vertex AI's own price list. Source: cloud.google.com/vertex-ai/generative-ai/pricing, collected 2026-11-21.

Google Vertex AI
Grok 4.6 (Xhigh) — benchmarks & cost | Benchmark Heaven