Tihu · Pelican Bench

Grok 4.5

xAI · Model release:

AI reference score 6.13 ± 0.72 Last 4 weekly runs · 12 images NormalUpdated every Sunday · 2026-10-26Top tiers of one model

Latest samples

MonLatest

Grok 4.5 · high · 2026-10-26 · #1 · 6.33Grok 4.5 · high · 2026-10-26 · #3 · 6.15Grok 4.5 · high · 2026-10-26 · #2 · 5.30

Score trend

high Normal

2026-10-08 · 6.482026-10-09 · 5.802026-10-10 · 6.082026-10-11 · 5.702026-10-12 · 6.402026-10-13 · 6.632026-10-14 · 6.502026-10-15 · 6.422026-10-16 · 6.932026-10-17 · 6.212026-10-18 · 5.692026-10-19 · 6.752026-10-20 · 6.542026-10-21 · 5.672026-10-22 · 6.852026-10-23 · 6.302026-10-24 · 6.712026-10-25 · 5.592026-10-26 · 5.93

19 records in the last 30 days

2026-10-26 · Normal2026-10-25 · Normal2026-10-24 · Normal2026-10-23 · Normal2026-10-22 · Normal2026-10-21 · Normal2026-10-20 · Normal2026-10-19 · Normal2026-10-18 · Normal2026-10-17 · Normal2026-10-16 · Normal2026-10-15 · Normal2026-10-14 · Normal2026-10-13 · Collecting earlier results2026-10-12 · Collecting earlier results2026-10-11 · Collecting earlier results2026-10-10 · Collecting earlier results2026-10-09 · Collecting earlier results2026-10-08 · Collecting earlier results

Earlier samples

high

Sun

Grok 4.5 · high · 2026-10-25 · #2 · 6.57Grok 4.5 · high · 2026-10-25 · #1 · 5.29Grok 4.5 · high · 2026-10-25 · #3 · 4.91

Sat

Grok 4.5 · high · 2026-10-24 · #2 · 6.86Grok 4.5 · high · 2026-10-24 · #1 · 6.74Grok 4.5 · high · 2026-10-24 · #3 · 6.54

Versions by score

  1. Grok 4.6
    Grok 4.6 · high
    high 6.79Last 4 weekly runs · 12 images
  2. Grok 4.5
    Grok 4.5 · high
    high 6.13Last 4 weekly runs · 12 images
  3. Grok 4.7
    Grok 4.7 · medium
    medium 6.09Last 4 weekly runs · 12 images Default tier for this version

Scores by criterion

Pelican likeness
5.83
Bicycle correctness
5.42
Riding pose
6.75
Composition
6.75
Obvious errors (higher is better)
5.00

Release notes

Original (Chinese)公开上市。官方 API changelog 记为 2026-07-08 在 xAI API 可用（grok-4.5）。此前有客户 beta。

· View release source ↗