Sonnet 5.5 vs GPT-6.1 Sol Six-Legged Walker Game

#godot #llm #gamedev #ai

Can an AI agent build a six-legged walker that never slides a foot? Claude Sonnet 5.5 in Claude Code and GPT-6.1 Sol in Codex got the same prompt and the same three reference pictures: a walker in Godot whose legs find the ground with raycasts, stay planted, step in a tripod gait and bend at the knee with inverse kinematics, over boulders, a trench and a flight of rock ledges. No keyframed walk, no downloaded or generated 3D models. Both worked directly in the Flockbay app, on the Godot engine, pressed Play themselves, and decided when they were done. Then each got one more message: make it look as realistic as possible.

Source: https://flockbay.com/blog/sonnet-5-5-vs-gpt-6-1-sol-six-legged-walker-in-godot/ · Updated 2026-10-06

The test

The two models cost the same per token. The bill, estimated at API prices, was not the same, and the video shows where it went.

Both walkers were driven by the same script on each game's own W, A, S and D.

The prompt

The exact prompt both models got:

Build a six-legged walker game in Godot. The player pilots a six-legged walking machine from a landing pad across a rocky canyon to a beacon on a ridge. W and S walk it forward and back, A and D turn it. On the way, the canyon floor is strewn with boulders, there is a trench to step across, and a flight of rock ledges to climb. The legs must be animated procedurally: each foot finds the ground with a raycast and stays planted while the body moves, then lifts and steps when it falls too far behind, and each leg is solved with inverse kinematics so the knee bends. Don't use a keyframed or pre-made walk animation, and don't let the feet slide. The body should ride over the terrain and tilt with it. Use a third-person camera behind the walker. Make the walker feel heavy, and make it look like the reference pictures. Build every model yourself from code or primitives; don't download or generate 3D models. Playtest it yourself and hand it over when you think it's done.

What was judged

How close it looks to the reference.

The walk: planted feet, reach, the gait, the knees, the body's ride.

Whether it finishes the course: the boulders, the trench, the ledges, the beacon.

What it cost: tokens, calls, time, and the estimated API bill.

The second pass: how real it looks, what that cost in tokens and in frame rate, and whether the walk survived it.

In the video

0:00 Opus won the last two
0:39 Six legs this time
1:20 The job
1:34 The models and the rules
2:07 Who drives
2:18 The reference
2:33 What a good walk looks like
3:16 What we are testing
3:37 Sonnet 5.5
4:25 Where Sonnet stops
5:06 GPT-6.1 Sol
6:14 Side by side
6:29 Neither stands on its legs
7:55 The second pass: make it real
9:14 What realism cost
9:35 The bill
10:18 Who takes it

It ran in Claude Code, on Flockbay’s MCP server

Claude Sonnet 5.5 ran in Claude Code for 28 minutes on the prompt above, with Flockbay’s tools connected over MCP: 143 calls to Flockbay’s tools and 20 to Claude Code’s own. Every call, from the run’s transcript:

Tool	Calls
`flockbay_script_write`	44
`flockbay_game_get`	28
`flockbay_game_input`	25
`flockbay_play`	21
`flockbay_game_set`	11
Bash	8
`flockbay_game_screenshot`	7
Read	7
Grep	4
`flockbay_script_errors`	3
`flockbay_last_run`	2
`flockbay_game_tree`	1
`flockbay_godot_status`	1
Write	1

No check failed along the way

Every check the run made passed. Its last, flockbay_script_errors, returned:

No errors in 6 scripts.
Scripts under res://addons were not checked; pass include_addons to check them.

res://scripts/walker.gd as it was handed over

Lines 1 to 40 of 665, from the game folder at the end of the build.

class_name Walker
extends CharacterBody3D
## The six-legged walking machine.
##
## Gait: every foot is planted at a world position and stays there. When a foot
## has fallen STEP_DIST behind its rest spot, its tripod (three alternating
## legs) lifts and swings it to a fresh foothold found with a raycast, while the
## other three carry the hull. Each leg is two-bone IK (see HexLeg). The hull
## rides a heavy spring above a plane fitted through the planted feet, so it
## follows and tilts with the ground. No animation clips.

signal footfall(pos: Vector3, strength: float)

const HIP_RISE := 3.2 # hull origin above the plane through the feet
const HULL_SCALE := 1.18
const MAX_SPEED := 3.6
const MAX_REVERSE := 2.3
const ACCEL := 1.5
const BRAKE := 2.6
const MAX_TURN := 0.62
const TURN_ACCEL := 1.8
const STEP_DIST := 1.25 # how far a foot may trail its rest spot before stepping
const HARD_DIST := 3.0 # beyond this the body must stop and wait for the legs
const SWING_TIME := 0.46
const DROP_MAX := 0.95 # foothold may be this far below the support plane
const RISE_MAX := 2.0 # ...or this far above it
const BODY_OMEGA := 6.0 # hull spring stiffness; low = heavy

var terrain: CanyonTerrain
var active := false
var control_enabled := true
var heading := 0.0
var speed := 0.0
var yaw_rate := 0.0
var input_forward := 0.0
var input_turn := 0.0
var speed_scale := 1.0
var body: Node3D
var hull_shape: Node3D
var legs: Array[HexLeg] = []

Download Flockbay free: Built in Flockbay, a free AI game maker built on Godot. https://flockbay.com/free-ai-game-maker/

I make Flockbay, a free AI game maker built on Godot. The original is on flockbay.com.

DEV Community