xAI Grok Statistics (2026)

Written by: Associate Editor
Published on:

Grok is xAI’s AI assistant, shipped across X (formerly Twitter) and grok.com, with versions that support real-time info, coding help, multimodal understanding, and image generation.

xAI Grok statistics
xAI Grok statistics

Key Grok stats at a glance

  • U.S. market share (Apptopia): 1.9% (Jan 2025) → 14.0% (Dec 2025) → 17.8% (Jan 2026).
  • Rank in the U.S.: Reported as the #3 most-used chatbot, behind ChatGPT and Google Gemini (Jan 2026).
  • Grok-2 benchmark highlights: MMLU 87.5%, HumanEval 88.4%, GPQA 56.0%, MATH 76.1%, DocVQA 93.6%.
  • Early model benchmarks: Grok-1 reported 73% on MMLU and 63.2% on HumanEval.
  • Early prototype scale: Grok-0 was trained with 33B parameters.
  • Grok 4.1 live preference (xAI): Preferred 64.78% of the time vs the previous production Grok model during a November 2025 rollout test.
  • LMArena Elo (xAI): Grok 4.1 Thinking reported at 1483 Elo (#1 overall); non-thinking mode reported at 1465 Elo (#2 overall).
  • “Grok in the Wild” (research preprint): Grok is called by millions weekly on X; it responds to 62% of requests; 51% of sampled interactions are in English; half of responses get ≤20 views after 48 hours (in their sample).

Grok usage and market share

Apptopia data reported by Reuters shows Grok’s U.S. market share rose from 1.9% in January 2025 to 17.8% in January 2026—about a 9.4× increase year over year.

Bar chart: Grok U.S. market share (Apptopia)

LabelBarValue
Jan 2025
1.9%
Dec 2025
14.0%
Jan 2026
17.8%

Max = 17.8%. Widths: Jan 2025 10.67%, Dec 2025 78.65%, Jan 2026 100.00%.

Model benchmark performance

xAI’s published Grok-2 benchmark table reports strong scores across general knowledge, math, and coding tasks.

  • MMLU: Grok-2 87.5% (vs Grok-1.5 81.3%).
  • HumanEval (pass@1): Grok-2 88.4% (vs Grok-1.5 74.1%).
  • MATH (maj@1): Grok-2 76.1% (vs Grok-1.5 50.6%).
  • GPQA: Grok-2 56.0% (vs Grok-1.5 35.9%).
  • DocVQA: Grok-2 93.6% (vs Grok-1.5 85.6%).

Bar chart: MMLU score comparison (from xAI’s Grok-2 benchmark table)

LabelBarValue
Grok-2
87.5%
GPT-4o
88.7%
Claude 3.5 Sonnet
88.3%
GPT-4 Turbo
86.5%

Max = 88.7%. Widths: Grok-2 98.65%, GPT-4o 100.00%, Claude 3.5 Sonnet 99.55%, GPT-4 Turbo 97.52%.

Pricing and access (X Premium tiers)

X Premium pricing varies by country and platform. Below are the published U.S. web prices; HUF figures use 1 USD = 319.23 HUF (official daily rate for February 17, 2026) and are approximate.

Tier (U.S. web)MonthlyAnnual (USD)Annual (HUF/year, approx.)
Basic$3.00$32.0010,215 HUF/year
Premium$8.00$84.0026,815 HUF/year
Premium+$40.00$395.00126,096 HUF/year

xAI funding milestones (context for Grok’s compute buildout)

Grok’s rapid iteration has been paired with major fundraising and infrastructure investment by xAI.

Bar chart: Selected xAI funding events

LabelBarValue
Series B (May 2024)
$6B
Series C (Dec 2024)
$6B
Debt+Equity (Jul 2025)
$10B
Series E (Jan 2026)
$20B

Max = $20B. Widths: Series B (May 2024) 30.00%, Series C (Dec 2024) 30.00%, Debt+Equity (Jul 2025) 50.00%, Series E (Jan 2026) 100.00%.

Sources

  • Reuters (Feb 13, 2026) — “Musk’s AI chatbot Grok gains US market share…” https://www.reuters.com/business/media-telecom/musks-ai-chatbot-groks-us-market-share-jumps-amid-sexualized-images-backlash-2026-02-13/
  • xAI — “Grok-2 Beta Release” (benchmark table) https://x.ai/news/grok-2
  • xAI — “Announcing Grok” (Grok-0 parameters; Grok-1 benchmark stats) https://x.ai/news/grok
  • xAI — “Grok 4.1” (preference rate; LMArena Elo) https://x.ai/news/grok-4-1
  • X Help Center — “Premium+ Price Adjustment” (pricing table) https://help.x.com/en/premium-plus-price-update
  • xAI — “Series B funding round” https://x.ai/news/series-b
  • xAI — “xAI raises $6B Series C” https://x.ai/news/series-c
  • Reuters (Jul 1, 2025) — “xAI raises $5 billion each in fresh debt and equity…” https://www.reuters.com/business/musks-xai-raises-5-billion-each-fresh-debt-equity-morgan-stanley-says-2025-07-01/
  • xAI — “xAI Raises $20B Series E” https://x.ai/news/series-e
  • Magyar Nemzeti Bank daily USD rate (Feb 17, 2026 shown via MNB-based listing) https://arfolyam.iridium.hu/en-US/USD
  • arXiv (Feb 11, 2026) — “Grok in the Wild: Characterizing the Roles and Uses of Large Language Models on Social Media” https://arxiv.org/abs/2602.11286