| Command R7B (12-2024) | Llama 4 Maverick | |
|---|---|---|
| Input, per 1M tokens | $0.0375 | $0.188 |
| Output, per 1M tokens | $0.15 | $0.652 |
| Context window | 128K | 1.05M |
| Classification or routing (600 tokens in, 10 out), per 1,000 | $0.024 | $0.119 |
| Chat reply (2,000 in, 400 out), per 1,000 | $0.135 | $0.636 |
| Retrieval-augmented answer (8,000 in, 500 out), per 1,000 | $0.375 | $1.83 |
| Confidence signal | no | yes |
| Tool calling | no | yes |
| BANKING77, correct of 400 | not measured | 294 |
Which one should carry your traffic?
If Command R7B (12-2024) answers your requests as well as Llama 4 Maverick, moving that traffic saves 79% on chat-shaped requests. Averages cannot tell you whether it does for your requests. Command R7B (12-2024) does not expose a confidence signal through OpenRouter, so Gimbix would test it as a full replacement rather than a first answer.
Command R7B (12-2024) profile · Llama 4 Maverick profile · Measure it on your traffic