---
title: "Runbook diffs | Pelican benchmark"
url: https://evaljetty.com/pelicans/runbooks.html
description: "Diff any two versions of the pelican-on-a-bicycle runbook across hand-curated and auto-evolved lineages."
updated: 2026-09-28
publisher: Jetty (https://jetty.io)
---

Runbook diffs · 98 runbooks · 14 lineages

# What changed between two runbooks

One lineage was edited by hand (Jon's claude-code + Sonnet sequence, v1 to v11). The rest were auto-evolved by the orchestrator: each round embeds the previous best SVG and names the weakest axis. Pick a lineage and any two versions.

Lineageclaude-code · Sonnet 4.6 · v1 · 11 runbooksclaude-code · Opus 4.7 · v1 · 10 runbooksgemini-cli · Gemini 3.5 Flash · v1 · 10 runbooksopencode · Gemini 3.5 Flash · v1 · 10 runbookshermes · Sonnet 4.6 · v1 · 10 runbookshermes · Gemini 3.5 Flash · v1 · 10 runbookshermes · GLM 5.1 · v1 · 10 runbookshermes · Fusion router · v1 · 3 runbooksopencode · Fusion router · v1 · 3 runbooksclaude-code · Sonnet 5 · v2 · 5 runbooksclaude-code · Opus 5.5 · v2 · 5 runbooksopencode · Gemini 3.8 Flash · v2 · 1 runbooksopencode · GPT-5.6 Terra · v2 · 5 runbookshermes · Jev Router · v2 · 5 runbooks
from
to
side by sideunified
 include baseline SVG

Loading…

## Hand-picked transitions

Click a card to load that pair.

---
Source: https://evaljetty.com/pelicans/runbooks.html · Evals by Jetty · Run your own evals: https://jetty.io/?utm_source=evaljetty&utm_medium=llms
