Tihu · Pelican Bench

GPT-6 Sol

Compare with previous version

OpenAI · Model release:

AI reference score 7.79 ± 0.65 · max (best) Last 4 weekly runs · 12 images
Tier comparison
TierSamplesAI reference scoreValid SVG rateMean timeSamples
max
GPT-6 Sol · max · 2026-10-26 · #3 · 8.04
 7.79 ± 0.65Last 4 weekly runs · 12 images 100%91 s3
medium default
GPT-6 Sol · medium · 2026-10-26 · #3 · 6.63
 6.29 ± 0.33Last 4 weekly runs · 12 images 100%41 s3
low
GPT-6 Sol · low · 2026-10-26 · #2 · 5.45
 5.45 ± 0.77Last 4 weekly runs · 12 images 100%26 s3
GPT-6 Sol · max · 2026-10-26 · #3 · 8.04
max 7.79 ± 0.65Last 4 weekly runs · 12 images
Valid SVG rate 100% · Mean time 91 s
GPT-6 Sol · medium · 2026-10-26 · #3 · 6.63
medium default6.29 ± 0.33Last 4 weekly runs · 12 images
Valid SVG rate 100% · Mean time 41 s
GPT-6 Sol · low · 2026-10-26 · #2 · 5.45
low 5.45 ± 0.77Last 4 weekly runs · 12 images
Valid SVG rate 100% · Mean time 26 s

Latest samples

max

AI reference score 7.79 ± 0.65 Last 4 weekly runs · 12 images Normal
Top tiers of one modelView tier page

MonLatest

GPT-6 Sol · max · 2026-10-26 · #3 · 8.04GPT-6 Sol · max · 2026-10-26 · #1 · 7.59GPT-6 Sol · max · 2026-10-26 · #2 · 7.18

medium

AI reference score 6.29 ± 0.33 Last 4 weekly runs · 12 images Normal
Top tiers of one modelView tier page

MonLatest

GPT-6 Sol · medium · 2026-10-26 · #3 · 6.63GPT-6 Sol · medium · 2026-10-26 · #2 · 6.38GPT-6 Sol · medium · 2026-10-26 · #1 · 5.94

low

AI reference score 5.45 ± 0.77 Last 4 weekly runs · 12 images Normal
Top tiers of one modelView tier page

MonLatest

GPT-6 Sol · low · 2026-10-26 · #2 · 5.45GPT-6 Sol · low · 2026-10-26 · #3 · 4.99GPT-6 Sol · low · 2026-10-26 · #1 · 4.16

Score trend

max Normal

2026-10-08 · 6.482026-10-09 · 6.832026-10-10 · 6.972026-10-11 · 8.152026-10-12 · 8.182026-10-13 · 7.992026-10-14 · 7.882026-10-15 · 7.892026-10-16 · 7.802026-10-17 · 7.632026-10-18 · 8.322026-10-19 · 7.882026-10-20 · 7.232026-10-21 · 7.522026-10-22 · 7.272026-10-23 · 7.402026-10-24 · 7.762026-10-25 · 8.402026-10-26 · 7.60

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

medium Normal

2026-10-08 · 6.952026-10-09 · 7.292026-10-10 · 6.552026-10-11 · 6.782026-10-12 · 6.612026-10-13 · 6.192026-10-14 · 6.752026-10-15 · 6.332026-10-16 · 6.132026-10-17 · 6.812026-10-18 · 5.982026-10-19 · 7.002026-10-20 · 6.822026-10-21 · 7.252026-10-22 · 6.722026-10-23 · 6.342026-10-24 · 6.252026-10-25 · 6.262026-10-26 · 6.31

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

low Normal

2026-10-08 · 5.862026-10-09 · 6.042026-10-10 · 6.732026-10-11 · 5.672026-10-12 · 5.402026-10-13 · 5.642026-10-14 · 5.752026-10-15 · 6.092026-10-16 · 5.192026-10-17 · 5.552026-10-18 · 6.572026-10-19 · 5.452026-10-20 · 5.192026-10-21 · 6.052026-10-22 · 6.602026-10-23 · 6.072026-10-24 · 5.342026-10-25 · 5.532026-10-26 · 4.87

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

max

medium

low

Versions by score

  1. GPT-6.1 Sol
    GPT-6.1 Sol · max
    max 8.58Last 4 weekly runs · 12 images
  2. GPT-6 Sol
    GPT-6 Sol · max
    max 7.79Last 4 weekly runs · 12 images
  3. GPT-5.6 Sol
    GPT-5.6 Sol · max
    max 7.71Last 4 weekly runs · 12 images

Scores by criterion

Scores by tier
Scoring criteria max medium low
Pelican likeness8.337.335.58
Bicycle correctness7.676.834.75
Riding pose6.504.674.17
Composition8.257.005.42
Obvious errors (higher is better)7.085.254.25

Release notes

Original (Chinese)与 GPT-6 Luna 同日发布。当天在 ChatGPT Work、Codex 和 API（gpt-6-sol）可用；公告写明 Chat 暂未上线。2026-10-07 起 GPT-6 Sol 进入 ChatGPT Chat（付费档）。

· View release source ↗