AI vs Puzzles

Back to Home
Puzzle#1722

Wordle

Released: 2026-03-07

7
Models Tested
3
Solved
5.86
Avg. Attempts
7
LLM Providers
Correct
Present
Absent

openai/gpt-4o

openai

Failed

t
a
b
l
e
b
r
i
d
e
c
h
o
k
e
s
o
l
v
e
w
r
o
t
e
f
r
o
z
e

mistralai/mistral-medium-3

mistral

Failed

c
r
a
n
e
m
o
o
s
e
h
o
n
e
y
l
o
d
g
e
b
o
g
u
e
r
o
g
u
e

google/gemini-2.0-flash-001

google

Failed

a
r
o
s
e
c
l
o
n
e
b
l
o
o
m
p
i
o
u
s
v
o
u
l
e
v
o
u
g
e

anthropic/claude-3.7-sonnet

anthropic

6/6

s
l
a
t
e
c
h
o
r
e
p
i
q
u
e
d
o
g
u
e
r
o
g
u
e
v
o
g
u
e

deepseek/deepseek-chat-v3-0324

deepseek

5/6

c
r
a
n
e
b
l
o
k
e
p
o
g
u
e
r
o
g
u
e
v
o
g
u
e

x-ai/grok-3-mini-beta

x-ai

6/6

c
r
a
n
e
f
l
u
t
e
m
o
u
s
e
r
o
u
t
e
b
o
u
l
e
v
o
g
u
e

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

meta

Failed

h
o
u
s
e
c
o
u
p
e
l
o
u
t
e
m
o
u
n
d
b
o
u
s
e
g
o
u
g
e

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