Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Composer 2 is clearly ahead on the aggregate, 62 to 49. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Composer 2's sharpest advantage is in agentic, where it averages 61.7 against 45. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 40% to 61.7%.
Composer 2 is the reasoning model in the pair, while Claude 3 Haiku is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use.
Pick Composer 2 if you want the stronger benchmark profile. Claude 3 Haiku only becomes the better choice if you would rather avoid the extra latency and token burn of a reasoning model.
Claude 3 Haiku
45
Composer 2
61.7
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Composer 2 is ahead overall, 62 to 49. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 40% and 61.7%.
Composer 2 has the edge for agentic tasks in this comparison, averaging 61.7 versus 45. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.