Can Claude Haiku 5.5 write working Godot 4 code? 14 of 15 on the first try
By Eduardo Orellana · Updated
Claude Haiku 5.5 passed 14 of 15 Godot 4 tasks on the first try and 15 of 15 after one fix with the test's output, each run headless on Godot 4.7.2. It missed: build and play an animation in code.
AppImage, x86_64. Ubuntu 22.04, Debian 12, Fedora 39 or newer.
Intel or AMD Chromebook, with Linux turned on.
Every model measured on the same Godot 4 tasks, each run headless on Godot 4.7.2 at a fixed 60 frames per second.
Measured by Flockbay: the same tasks, the same checks, one fix allowed with the test output.
| Model | First try | After one fix | Agent | Run on | Read from |
|---|---|---|---|---|---|
| Claude Opus 5.5 | 15 of 15 | 15 of 15 | Claude Code | 2026-10-06 | Claude Opus 5.5 results |
| Claude Sonnet 5.5 | 15 of 15 | 15 of 15 | Claude Code | 2026-10-06 | Claude Sonnet 5.5 results |
| GPT-6.1 Sol | 15 of 15 | 15 of 15 | Codex | 2026-10-06 | GPT-6.1 Sol results |
| DeepSeek V4.1 Flash | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | DeepSeek V4.1 Flash results |
| Ember-1 | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | Ember-1 results |
| Gemini 3.1 Pro Preview | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | Gemini 3.1 Pro Preview results |
| Muse Wonder | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | Muse Wonder results |
| Kimi K3 | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | Kimi K3 results |
| GPT-5.6 Luna | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | GPT-5.6 Luna results |
| GPT-6 Luna | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | GPT-6 Luna results |
| Hy4 preview | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | Hy4 preview results |
| Pareto 26.10 | 15 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | Pareto 26.10 results |
| Claude Haiku 5.5 | 14 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | Claude Haiku 5.5 results |
| MiMo-V2.6-Pro-UltraSpeed | 14 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | MiMo-V2.6-Pro-UltraSpeed results |
| GLM 5.2 | 14 of 15 | 15 of 15 | Hosted, by API | 2026-10-18 | GLM 5.2 results |
| DeepSeek V4 Pro 0813 | 14 of 15 | 14 of 15 | Hosted, by API | 2026-10-18 | DeepSeek V4 Pro 0813 results |
| GPT-5.4 | 14 of 15 | 14 of 15 | Hosted, by API | 2026-10-18 | GPT-5.4 results |
| MiniMax M3 | 13 of 15 | 14 of 15 | Hosted, by API | 2026-10-18 | MiniMax M3 results |
| Qwen3.7 Max | 13 of 15 | 14 of 15 | Hosted, by API | 2026-10-18 | Qwen3.7 Max results |
| Fugu Max | 13 of 15 | 14 of 15 | Hosted, by API | 2026-10-18 | Fugu Max results |
| Beam 1.0 501B A23B | 12 of 15 | 14 of 15 | Hosted, by API | 2026-10-18 | Beam 1.0 501B A23B results |
| Mistral Large 4 | 12 of 15 | 13 of 15 | Hosted, by API | 2026-10-18 | Mistral Large 4 results |
| Claude Haiku 4.5 | 11 of 15 | 14 of 15 | Claude Code | 2026-10-06 | Claude Haiku 4.5 results |
| Nemotron 3.5 Super VL | 11 of 15 | 14 of 15 | Hosted, by API | 2026-10-18 | Nemotron 3.5 Super VL results |
| LongCat 2.0 | 11 of 15 | 13 of 15 | Hosted, by API | 2026-10-18 | LongCat 2.0 results |
| Grok 4.3 | 11 of 15 | 13 of 15 | Hosted, by API | 2026-10-18 | Grok 4.3 results |
| Laguna S 2.1 | 9 of 15 | 12 of 15 | Hosted, by API | 2026-10-18 | Laguna S 2.1 results |
| Nova 2 Lite | 2 of 15 | 4 of 15 | Hosted, by API | 2026-10-18 | Nova 2 Lite results |
How the test works
Each task asks for one GDScript file with a stated API, in a fresh Godot project. Claude Haiku 5.5 was asked through its API, the way Flockbay reaches its hosted models: the task in one request, no tools, and the whole file back in a code block. A check scene then loaded the file and drove it on Godot 4.7.2, headless, at a fixed 60 frames per second: it pressed input actions, stepped physics, waited on timers and signals, and compared the results.
When a task failed, the model got the check's output once and could fix its file. Every task touches an API that changed between Godot 3 and Godot 4, or was added since: move_and_slide and is_on_floor, FileAccess and JSON, signals, tweens, Timer nodes, PackedScene, TileMapLayer, ray queries, AStarGrid2D, AnimationLibrary, and lambdas, which capture local variables by value.
The checks themselves are tested first: each passes on a reference answer and fails on an empty script. Run on 2026-10-18.
Claude Haiku 5.5 in Flockbay
Claude Haiku 5.5 is a hosted model in the Flockbay app's model menu, on the Indie Developer plan and above. It uses the plan's hosted usage about 10.7 times as fast as Auto.
It takes images, so Flockbay can show it screenshots of your game, and it has an effort setting in the composer, from quick answers to longer reasoning.
Move a CharacterBody2D with an input action: pass
The task: Write res://player.gd for Godot 4: a script for a CharacterBody2D that moves right at 200 pixels per second while the input action "ui_right" is held, and stands still otherwise.
Passed on the first try.
extends CharacterBody2D
const SPEED: float = 200.0
func _physics_process(_delta: float) -> void:
if Input.is_action_pressed("ui_right"):
velocity.x = SPEED
else:
velocity.x = 0.0
velocity.y = 0.0
move_and_slide()
Fall with gravity and land on the floor: pass
The task: Write res://ball.gd for Godot 4: a script for a CharacterBody2D that falls with a gravity of 980 pixels per second squared and stops when it lands on the floor. The scene gives it its collision shape.
Passed on the first try.
extends CharacterBody2D
const GRAVITY := 980.0
func _physics_process(delta: float) -> void:
if is_on_floor():
velocity.y = 0.0
else:
velocity.y += GRAVITY * delta
move_and_slide()
A score label: pass
The task: Write res://score_label.gd for Godot 4: a script for a Label that shows "Score: 0" when it enters the scene, and has a method add(points: int) that adds to the score and updates the text to "Score: <total>".
Passed on the first try.
extends Label
var score: int = 0
func _ready() -> void:
text = "Score: %d" % score
func add(points: int) -> void:
score += points
text = "Score: %d" % score
Save and load a dictionary as JSON: pass
The task: Write res://save_game.gd for Godot 4: a script extending RefCounted with two static functions: save_data(data: Dictionary, path: String) -> void writes the dictionary to the file as JSON, and load_data(path: String) -> Dictionary reads it back.
Passed on the first try.
extends RefCounted
static func save_data(data: Dictionary, path: String) -> void:
var file := FileAccess.open(path, FileAccess.WRITE)
if file == null:
push_error("save_game: could not open '%s' for writing (error %d)" % [path, FileAccess.get_open_error()])
return
file.store_string(JSON.stringify(data))
file.close()
static func load_data(path: String) -> Dictionary:
if not FileAccess.file_exists(path):
push_warning("save_game: file '%s' does not exist" % path)
return {}
var file := FileAccess.open(path, FileAccess.READ)
if file == null:
push_error("save_game: could not open '%s' for reading (error %d)" % [path, FileAccess.get_open_error()])
return {}
var text := file.get_as_text()
file.close()
var parsed = JSON.parse_string(text)
if typeof(parsed) != TYPE_DICTIONARY:
push_error("save_game: '%s' does not contain a valid JSON dictionary" % path)
return {}
return parsed
Health with signals: pass
The task: Write res://health.gd for Godot 4: a script extending Node with health starting at 100, a signal health_changed(old_value: int, new_value: int), a signal died, and a method take_damage(amount: int) that lowers health without going below 0, emits health_changed when health changes, and emits died once, when health reaches 0.
Passed on the first try.
extends Node
signal health_changed(old_value: int, new_value: int)
signal died
var health: int = 100
func take_damage(amount: int) -> void:
if amount <= 0 or health <= 0:
return
var old_value := health
health = maxi(health - amount, 0)
if health != old_value:
health_changed.emit(old_value, health)
if old_value > 0 and health == 0:
died.emit()
Fade out with a tween: pass
The task: Write res://fader.gd for Godot 4: a script extending Node2D with a method fade_out(duration: float) that fades the node's modulate alpha from 1 to 0 over duration seconds with a tween, and a signal faded emitted when the fade ends.
Passed on the first try.
extends Node2D
signal faded
func fade_out(duration: float) -> void:
var tween := create_tween()
tween.tween_property(self, "modulate:a", 0.0, duration).from(1.0)
tween.finished.connect(func() -> void: faded.emit())
Spawn on a timer, then stop: pass
The task: Write res://spawner.gd for Godot 4: a script extending Node that, once in the scene, adds a new child Node2D every 0.25 seconds, using a Timer, and stops after 4 children. It keeps the number spawned in a variable spawned_count.
Passed on the first try.
extends Node
const MAX_SPAWNS: int = 4
const SPAWN_INTERVAL: float = 0.25
var spawned_count: int = 0
var timer: Timer
func _ready() -> void:
timer = Timer.new()
timer.wait_time = SPAWN_INTERVAL
timer.one_shot = false
timer.timeout.connect(_on_timer_timeout)
add_child(timer)
timer.start()
func _on_timer_timeout() -> void:
if spawned_count >= MAX_SPAWNS:
timer.stop()
return
var node := Node2D.new()
node.name = "Spawned_%d" % (spawned_count + 1)
add_child(node)
spawned_count += 1
if spawned_count >= MAX_SPAWNS:
timer.stop()
A door with states: pass
The task: Write res://door.gd for Godot 4: a script extending Node for a door with the states CLOSED, OPENING and OPEN. It starts CLOSED. A method interact() starts opening a closed door; 0.5 seconds of game time later it is OPEN. A method get_state_name() -> String returns "CLOSED", "OPENING" or "OPEN".
Passed on the first try.
extends Node
enum State { CLOSED, OPENING, OPEN }
const OPEN_DURATION: float = 0.5
var state: State = State.CLOSED
var _elapsed: float = 0.0
func _ready() -> void:
set_process(false)
func _process(delta: float) -> void:
if state != State.OPENING:
set_process(false)
return
_elapsed += delta
if _elapsed >= OPEN_DURATION:
_elapsed = 0.0
state = State.OPEN
set_process(false)
func interact() -> void:
if state != State.CLOSED:
return
state = State.OPENING
_elapsed = 0.0
set_process(true)
func get_state_name() -> String:
match state:
State.CLOSED:
return "CLOSED"
State.OPENING:
return "OPENING"
State.OPEN:
return "OPEN"
return "UNKNOWN"
An inventory with counts: pass
The task: Write res://inventory.gd for Godot 4: a script extending RefCounted with add_item(item_name: String, count: int), remove_item(item_name: String, count: int) -> bool that removes only when there are enough and returns whether it did, and get_count(item_name: String) -> int that is 0 for an item never added.
Passed on the first try.
extends RefCounted
var _items: Dictionary = {}
func add_item(item_name: String, count: int) -> void:
if count <= 0:
return
_items[item_name] = get_count(item_name) + count
func remove_item(item_name: String, count: int) -> bool:
if count <= 0:
return false
var current: int = get_count(item_name)
if current < count:
return false
var remaining: int = current - count
if remaining == 0:
_items.erase(item_name)
else:
_items[item_name] = remaining
return true
func get_count(item_name: String) -> int:
return _items.get(item_name, 0)
Load a level scene from a path: pass
The task: Write res://level_loader.gd for Godot 4: a script extending Node with a method load_level(path: String) -> Node that loads the scene file at path, creates an instance of it, adds it as a child, and returns it.
Passed on the first try.
extends Node
func load_level(path: String) -> Node:
var scene := load(path) as PackedScene
if scene == null:
push_error("LevelLoader: failed to load scene at '%s'" % path)
return null
var instance := scene.instantiate()
if instance == null:
push_error("LevelLoader: failed to instantiate scene at '%s'" % path)
return null
add_child(instance)
return instance
Paint cells on a TileMapLayer: pass
The task: Write res://grid_painter.gd for Godot 4.7: a script extending TileMapLayer with paint_row(y: int, from_x: int, to_x: int) that sets the cells from (from_x, y) to (to_x, y) inclusive to tile source 0, atlas coordinates (0, 0), and is_painted(cell: Vector2i) -> bool that says whether a cell has a tile.
Passed on the first try.
extends TileMapLayer
const PAINT_SOURCE_ID: int = 0
const PAINT_ATLAS_COORDS: Vector2i = Vector2i(0, 0)
## Paints cells from (from_x, y) to (to_x, y), inclusive, with source 0 and atlas (0, 0).
## If from_x is greater than to_x, the range is still painted.
func paint_row(y: int, from_x: int, to_x: int) -> void:
var start_x: int = mini(from_x, to_x)
var end_x: int = maxi(from_x, to_x)
for x in range(start_x, end_x + 1):
set_cell(Vector2i(x, y), PAINT_SOURCE_ID, PAINT_ATLAS_COORDS)
## Returns true if the cell currently has a tile.
func is_painted(cell: Vector2i) -> bool:
return get_cell_source_id(cell) != -1
Cast a physics ray: pass
The task: Write res://line_of_sight.gd for Godot 4: a script extending Node2D with first_hit(from: Vector2, to: Vector2) -> Vector2 that casts a physics ray through the 2D world from `from` to `to` and returns the point where it first hits a collider, or `to` when it hits nothing. It is called from physics processing.
Passed on the first try.
extends Node2D
# Casts a physics ray from `from` to `to` through the 2D world.
# Returns the point of the first collider hit, or `to` if nothing is hit.
# Must be called during physics processing (_physics_process), since
# direct_space_state is only valid during the physics step.
func first_hit(from: Vector2, to: Vector2) -> Vector2:
var world := get_world_2d()
if world == null:
return to
var space_state := world.direct_space_state
var query := PhysicsRayQueryParameters2D.create(from, to)
var result := space_state.intersect_ray(query)
if result.is_empty():
return to
return result["position"]
A path on a grid around walls: pass
The task: Write res://path_finder.gd for Godot 4: a script extending RefCounted with find_path(size: Vector2i, walls: Array[Vector2i], from: Vector2i, to: Vector2i) -> Array[Vector2i] that returns the shortest path of grid cells from `from` to `to`, both included, on a grid of `size` cells, never entering a wall cell and moving only up, down, left or right.
Passed on the first try.
extends RefCounted
# Returns the shortest 4-directional path of grid cells from `from` to `to`
# (both included) on a grid of `size` cells, never entering a wall cell.
# Returns an empty array if no path exists or the endpoints are invalid.
@warning_ignore("integer_division")
func find_path(size: Vector2i, walls: Array[Vector2i], from: Vector2i, to: Vector2i) -> Array[Vector2i]:
var result: Array[Vector2i] = []
if size.x <= 0 or size.y <= 0:
return result
if not _in_bounds(size, from) or not _in_bounds(size, to):
return result
var blocked := {}
for w in walls:
blocked[w] = true
if blocked.has(from) or blocked.has(to):
return result
if from == to:
result.append(from)
return result
var total := size.x * size.y
var parent := PackedInt32Array()
parent.resize(total)
parent.fill(-1)
var start_i := _idx(size, from)
var goal_i := _idx(size, to)
parent[start_i] = start_i
var queue := PackedInt32Array([start_i])
var head := 0
var found := false
var dirs: Array[Vector2i] = [
Vector2i(1, 0),
Vector2i(-1, 0),
Vector2i(0, 1),
Vector2i(0, -1),
]
while head < queue.size() and not found:
var cur := queue[head]
head += 1
var cp := Vector2i(cur % size.x, cur / size.x)
for d in dirs:
var np: Vector2i = cp + d
if not _in_bounds(size, np) or blocked.has(np):
continue
var ni := _idx(size, np)
if parent[ni] != -1:
continue
parent[ni] = cur
if ni == goal_i:
found = true
break
queue.append(ni)
if not found:
return result
var idx := goal_i
while idx != start_i:
result.push_front(Vector2i(idx % size.x, idx / size.x))
idx = parent[idx]
result.push_front(from)
return result
func _in_bounds(size: Vector2i, p: Vector2i) -> bool:
return p.x >= 0 and p.y >= 0 and p.x < size.x and p.y < size.y
func _idx(size: Vector2i, p: Vector2i) -> int:
return p.y * size.x + p.x
Build and play an animation in code: pass after one fix
The task: Write res://blinker.gd for Godot 4: a script extending Node2D that, when it enters the scene, creates an AnimationPlayer child and a looping animation named "blink", 1 second long, that keeps the node visible for the first half of each second and hidden for the second half, and plays it.
Failed on the first try (BENCH FAIL: visible at 0.25, 0.75, 1.25 and 1.75 s was [true, true, true, true], expected [true, false, true, false]), passed after one fix.
What Godot 4.7.2 printed on the first try:
More on this: Can Claude Haiku 4.5 write working Godot 4 code? 11 of 15 on the first try
Godot Engine v4.7.2.stable.mono.official.ed1daf0bf - https://godotengine.org
SCRIPT ERROR: Invalid call. Nonexistent function 'track_set_update_mode' in base 'Animation'.
at: _ready (res://blinker.gd:21)
GDScript backtrace (most recent call first):
[0] _ready (res://blinker.gd:21)
[1] _ready (res://bench_check.gd:10)
BENCH FAIL: visible at 0.25, 0.75, 1.25 and 1.75 s was [true, true, true, true], expected [true, false, true, false]Wait for a signal or a timeout: pass
The task: Write res://waiter.gd for Godot 4: a script extending Node with wait_for(sig: Signal, seconds: float) -> bool, a coroutine that waits until the signal is emitted or `seconds` of game time pass, whichever comes first, and returns true if the signal came first. The signal has no arguments.
Passed on the first try.
extends Node
## Waits until `sig` is emitted or `seconds` of game time pass, whichever comes first.
## Returns true if the signal was emitted first, false if the timeout came first.
func wait_for(sig: Signal, seconds: float) -> bool:
# A Dictionary is shared by reference with the lambdas, so they can update it.
var state := {"done": false, "signal_first": false}
var timer := get_tree().create_timer(seconds)
var on_signal := func() -> void:
if state["done"]:
return
state["done"] = true
state["signal_first"] = true
var on_timeout := func() -> void:
if state["done"]:
return
state["done"] = true
sig.connect(on_signal, CONNECT_ONE_SHOT)
timer.timeout.connect(on_timeout, CONNECT_ONE_SHOT)
while not state["done"]:
await get_tree().process_frame
# Clean up whichever connection did not fire.
if sig.is_connected(on_signal):
sig.disconnect(on_signal)
if timer.timeout.is_connected(on_timeout):
timer.timeout.disconnect(on_timeout)
return state["signal_first"]
Questions
- Is Claude Haiku 5.5 good at Godot?
- On these 15 tasks it passed 14 on the first try and 15 after one fix. Each task is small and checked by running it, so the score says whether its Godot 4 code works, not how well it designs a game.
Related
- Claude Opus 5.5 on the same tasks
- Claude Sonnet 5.5 on the same tasks
- DeepSeek V4.1 Flash on the same tasks
- Ember-1 on the same tasks
- Gemini 3.1 Pro Preview on the same tasks
- GLM 5.2 on the same tasks
- GPT-5.6 Luna on the same tasks
- GPT-6 Luna on the same tasks
- Every verified Godot 4 answer
- Can MiniMax M3 write working Godot 4 code? 13 of 15 on the first try
- Can Beam 1.0 501B A23B write working Godot 4 code? 12 of 15 on the first try
- Can Muse Wonder write working Godot 4 code? 15 of 15 on the first try
- Can Claude Haiku 4.5 write working Godot 4 code? 11 of 15 on the first try
- Can GPT-5.4 write working Godot 4 code? 14 of 15 on the first try
Flockbay runs your own agent, Claude or ChatGPT on the official Godot 4.7.2, and gives it Godot's errors the way this test does.
Download Flockbay free