Side-by-side technical and economic comparison between Alibaba / Qwen's Qwen 2.5 72B Instruct — Together AI and Cohere's Command R (08-2024).
• Qwen 2.5 72B Instruct — Together AI: Classified as a Legacy Reference model (Deprecated on Together AI (Reference)).
• Command R (08-2024): Classified as a Legacy Reference model (Tracked Production / Legacy Family).
| Metric | Qwen 2.5 72B Instruct — Together AI | Command R (08-2024) | Advantage |
|---|---|---|---|
| Provider Organization | Alibaba / Qwen | Cohere | — |
| Standard Input / 1M | $0.35 | $0.15 | Command R (08-2024) (57% lower) |
| Cached Input / 1M | $0.175 | $0.015 | Command R (08-2024) lower |
| Output / 1M | $0.40 | $0.60 | Qwen 2.5 72B Instruct — Together AI lower |
| Context Window | 128k | 128k | Qwen 2.5 72B Instruct — Together AI (128k) |
| Max Generation Tokens | 8.2k | 4.1k | Qwen 2.5 72B Instruct — Together AI |
| Tokenizer Family | Qwen Byte-level BPE (~152k vocabulary) | Cohere BPE (~256k vocabulary) | — |
Opt for Qwen 2.5 72B Instruct — Together AI when your engineering requirements prioritize Alibaba / Qwen's ecosystem, specific tokenizer efficiencies (Calibrated Qwen Tokenizer (±3%)), or when your expected prompt-to-completion ratios favor its $0.35/M input rate.
Opt for Command R (08-2024) when looking for Cohere's tooling integration, specific context window depth (128k tokens), or when output generation volume favors its $0.6/M rate.