| Command R+ | GPT-4o | |
|---|---|---|
| arenaElo | 1213 | 1345 |
| humanEval | 71.5 | 90.2 |
| sweBench | 20 | 33 |
| mmlu | 75.7 | 88.7 |
| gpqa | 42 | 53.6 |
| aime | 25 | 42 |
| Price in/out ($/M) | $2.5 / $10 | $2.5 / $10 |
These two are unusually easy to compare because the commercial terms are identical: $2.5 / $10 per M tokens and a 128K context window on both. Everything else favours GPT-4o, which leads on every metric the two publish, among them MMLU 88.7% against 75.7%, HumanEval 90.2% against 71.5%, SWE-bench Verified 33% against 20% and 1345 Arena Elo against 1213, and it adds audio on top of text and vision where Command R+ is text only. The more useful conclusion is the one neither vendor will print: both are 2024 models being sold at 2024 prices, and in 2026 that same $2.5 buys far more elsewhere. If you are picking between these two today, the real answer is that your upgrade path matters more than this duel.