ModelCap
Current AI model comparison
Exact identities · public evidence
OpenAI
GPT-3.5 Turbo Instruct
Position
#108
Index
22.1
Estimated
publisher corpus prior · leave-one-anchor-out calibrated
Comparison side 1
OpenAI
o3
Position
#37
Index
55.1
Measured
4 public benchmark observations across 4 boards
Comparison side 2
Output / 1M
$2.00
vs $8.00
Context
4K
vs 200K
Weights
API only
vs API only