AI vs Puzzles

Back to Home
Puzzle#1702

Wordle

Released: 2026-02-15

7
Models Tested
1
Solved
5.86
Avg. Attempts
7
LLM Providers
Correct
Present
Absent

mistralai/mistral-medium-3

mistral

Failed

c
r
a
n
e
d
o
l
l
y
p
i
l
o
t
b
i
l
l
y
h
e
l
l
s
s
w
i
l
l

google/gemini-2.0-flash-001

google

Failed

a
r
o
s
e
l
u
c
k
y
b
l
u
s
h
f
l
o
s
s
c
l
u
m
p
s
l
u
r
p

openai/gpt-4o

openai

Failed

l
i
g
h
t
l
o
v
e
s
s
a
l
o
n
s
l
u
m
p
s
l
u
s
h
s
k
u
l
k

anthropic/claude-3.7-sonnet

anthropic

Failed

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

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

meta

Failed

h
o
u
s
e
r
u
s
t
y
l
u
n
c
h
s
l
u
s
h
s
l
u
m
s
s
l
a
f
t

deepseek/deepseek-chat-v3-0324

deepseek

Failed

c
r
a
n
e
s
l
i
m
y
s
l
o
t
h
s
l
e
e
k
s
l
u
n
k
s
l
u
r
p

x-ai/grok-3-mini-beta

x-ai

5/6

s
l
a
t
e
s
i
l
l
y
s
o
u
l
s
s
c
u
l
l
s
k
u
l
l

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