Lab Comparison
Last updated , from board data generated at 17:52 UTC.
All 3 ranked Cohere configurations on ToneBench, side by side: Writing Elo, Writing Score, the nine metrics, and cost and time per script. A configuration is one model at one thinking level.
Ranks in these notes are configuration ranks: each setting's place among all 151 configurations. The Highest Writing Elo card above gives the model rank.
Picks and Notes From Louis
If you have to use Cohere, use Cohere Command A+ (05-2026) (thinking on). It's the best writer here, and on the route the benchmark used, each script came out free: Cohere serves it at no charge up to its rate limits. That's a free tier, not a list price, so don't count on it lasting. Cohere Command A (03-2025) (no thinking) is last in the family and the only one with a price on that route, at $0.046 a script, so I wouldn't use it for scripts at all.
Free or not, I wouldn't pick Cohere for writing. The family's best sits far behind cheap open models from other labs.
All 3 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) (thinking on), is at #142, and the family drops another 261 Elo from there, down to Cohere Command A (03-2025) (no thinking) at #146.
Every row wrote the same 10 scripts for the same four blind judges, so these Writing Elo ratings sit on the leaderboard's scale. The rank column is the configuration rank, each setting's place among all 151 configurations on the board. Each row is a different model. The ± after each Elo is half the width of its 95% range, and scores run from 0 to 100. Cost is one full script at the API list price of the route used. A row marked free ran on a route that served it at no charge, so its cost is $0. Each name opens its scorecard, and the head-to-head comparisons cover models from different labs.
With the metric scores shown, the table scrolls sideways.
| Configuration Rank | Configuration | Writing Elo | Writing Score | Cost / Script | Time / Script | Tone & Voice Match | Writing Craft & Clarity | Substance, Accuracy & Value | Hook Strength | YouTube Best Practices | Continuity & Emotion | Slop Score | Length Adherence | Visual Cue Quality |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| #142 | Cohere Command A+ (05-2026) (thinking on) · open weights | 363 ±62 | 56.7 | free | 1.6 min | 48.5 | 57.8 | 56.8 | 69.9 | 56.3 | 51.5 | 72.3 | 59.0 | 54.5 |
| #143 | Cohere Command A Reasoning (thinking on) · open weights | 215 ±53 | 52.7 | free | 0.4 min | 48.1 | 50.2 | 56.1 | 71.1 | 52.9 | 56.1 | 62.8 | 25.6 | 54.1 |
| #146 | Cohere Command A (03-2025) (no thinking) · open weights | 102 ±38 | 49.7 | $0.046 | 0.5 min | 43.6 | 52.0 | 52.4 | 67.2 | 48.6 | 50.6 | 69.2 | 19.4 | 53.0 |
Each comparison lists every thinking level of both models, with Writing Elo, cost and time side by side. The line under each pair gives the Writing Elo gap between the two best configurations. A pair outside the written match-ups opens in the comparison tool.