3 ms·
Why is this article based on a chart that has GPT-5.5 and Claude Opus 4.8 as the newest evaluated models? Those are very old now. There have been massive impro
by heaney-555 3mo ago
Why is this article based on a chart that has GPT-5.5 and Claude Opus 4.8 as the newest evaluated models? Those are very old now.
There have been massive improvements in computer use in GPT 5.6 and Claude 5.
- mpavlov 3mo agoUnfortunately, mainly because there're only self-reported and partial numbers for the benchmarks that matter the most.