Every Cohere model, side by side
All 6 Cohere configurations rank near the bottom of the board for script writing, and that isn't really a knock on Cohere. It builds its Command models mostly for enterprise work like retrieval and tool use, and they may well be good at that job. Writing a YouTube script in someone else's voice is a different job. The best of them, Cohere Command A+ (05-2026), is at #167, and the family drops another 845 Elo from there, down to Cohere Command R7B at #181 with an Elo of -359.
All 6 configurations
Every configuration here wrote the same 10 scripts for the same four blind judges, so the writing Elo and ranks compare across the whole board, not just inside this family. A highlighted row is a model's best configuration, and the rows under it are the same model at other effort or thinking settings. Scores run 0 to 100, and cost is one full script at the API list price of the route we ran.
| Board rank | Configuration | Elo | Overall | Tone | Craft | Substance | Hook | YouTube | Flow | Slop | Length | Cues | Cost / script | Time / script |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| #167 | Cohere Command A+ (05-2026) · open | 486 ±63 | 63.0 | 63.9 | 64.1 | 59.8 | 72.7 | 63.2 | 55.3 | 75.9 | 58.0 | 64.3 | free | 0.9 min |
| #168 | Cohere Command A Reasoning · open | 436 ±37 | 61.8 | 61.7 | 66.5 | 63.1 | 78.2 | 58.9 | 63.1 | 67.2 | 32.5 | 57.3 | free | 0.4 min |
| #171 | Cohere Command A (03-2025) · open | 308 ±39 | 59.1 | 58.0 | 65.2 | 61.5 | 71.4 | 57.0 | 58.6 | 73.2 | 25.6 | 63.3 | $0.042 | 0.3 min |
| #176 | Cohere Command R · open | 1 ±43 | 51.4 | 50.4 | 60.3 | 58.9 | 56.6 | 42.6 | 51.3 | 73.1 | 23.9 | 41.2 | $0.0025 | 3.2 min |
| #177 | Cohere Command R+ · open | -106 ±78 | 42.0 | 40.1 | 47.1 | 47.5 | 47.0 | 35.1 | 38.6 | 73.8 | 17.0 | 44.4 | $0.049 | 4.1 min |
| #181 | Cohere Command R7B · open | -359 ±63 | 31.1 | 28.1 | 30.3 | 36.9 | 32.2 | 24.5 | 22.9 | 69.5 | 31.3 | 22.0 | $0.0008 | 0.5 min |
Which one should you use?
If you have to use Cohere, use Cohere Command A+ (05-2026). It's the best writer here and the cheapest too: on the route we ran it through, each script came out free. That's a fact about that route, not a price you can count on everywhere. And it's well ahead of the older Command R models, so the newer Command A line is the only part of this family I'd consider. Cohere Command R7B is last in the family, and I wouldn't use it for scripts at all.
Honestly, though, I wouldn't pick Cohere for writing. Even the family's best sits far behind cheap open models from other labs.
Every row links to its full scorecard, and the cross-lab match-ups live in the head-to-head comparisons.
← All comparisons