AI vs Puzzles

Back to Home
Puzzle#1709

Wordle

Released: 2026-02-22

7
Models Tested
3
Solved
5.57
Avg. Attempts
7
LLM Providers
Correct
Present
Absent

openai/gpt-4o

openai

5/6

c
r
a
n
e
p
l
a
i
n
t
h
a
w
s
b
e
l
a
y
g
u
a
v
a

anthropic/claude-3.7-sonnet

anthropic

4/6

s
l
a
t
e
b
r
a
i
n
q
u
a
c
k
g
u
a
v
a

mistralai/mistral-medium-3

mistral

Failed

c
r
a
n
e
d
i
a
r
y
s
t
a
m
p
b
a
k
e
r
b
l
a
z
e
w
a
c
k
y

meta-llama/llama-3.3-70b-instruct

meta

Failed

h
o
u
s
e
b
r
u
n
t
i
c
i
n
a
l
u
x
u
a
b
u
o
y
a
m
u
d
r
a

deepseek/deepseek-chat-v3-0324

deepseek

Failed

c
r
a
n
e
d
e
b
u
t
q
u
i
l
t
h
u
m
o
r
p
u
d
g
y
f
u
g
u
e

x-ai/grok-3-mini-beta

x-ai

6/6

s
l
a
t
e
c
h
a
i
n
b
o
a
r
d
q
u
a
f
f
q
u
a
l
m
g
u
a
v
a

google/gemini-2.0-flash-001

google

Failed

c
r
a
n
e
b
l
a
s
t
d
a
i
s
y
u
m
a
j
o
a
v
a
s
t
v
a
p
i
d

Methodology Note

Each model receives the same prompt structure with Wordle rules and feedback format. Models are given a maximum of 6 attempts with standard positional feedback after each guess to help refine their next attempt.

Leaderboard