Cohere (5 models) vs DeepSeek (11 models) - compared across composite scores, pricing, capabilities, and context windows.
| Cohere | Score | vs | DeepSeek | Score |
|---|---|---|---|---|
| Command A | 50 | R1 Distill Llama 70B | 41 | |
| Command R+ (08-2024) | 48 | DeepSeek V3.1 Terminus | 69 | |
| Command R (08-2024) | 48 | DeepSeek V3 | 69 | |
| North Mini Code (free) | 40 | DeepSeek V3.1 | 69 | |
| Command R7B (12-2024) | 35 | DeepSeek V3.2 Exp | 70 |
| Capability | Cohere | DeepSeek | Leader |
|---|---|---|---|
Vision | 0/5 | 0/11 | Tie |
Reasoning | 1/5 | 9/11 | DeepSeek |
Function Calling | 3/5 | 10/11 | DeepSeek |
JSON Mode | 4/5 | 10/11 | DeepSeek |
Web Search | 0/5 | 0/11 | Tie |
Streaming | 5/5 | 11/11 | DeepSeek |
Image Output | 0/5 | 0/11 | Tie |
| Metric | Cohere | DeepSeek |
|---|---|---|
| Cheapest Input (per 1M tokens) | $0.037 Command R7B (12-2024) | $0.090 DeepSeek V4 Flash |
| Cheapest Output (per 1M tokens) | $0.150 | $0.180 |
| Most Expensive Input (per 1M tokens) | $2.50 Command A | $0.800 R1 |
| Most Expensive Output (per 1M tokens) | $10.00 | $2.50 |
| Free Models | 1 | 0 |
| Max Context Window | 256K | 1.0M |
| Model | Score | Input $/M | Output $/M |
|---|---|---|---|
| Command A | 50 | $2.50 | $10.00 |
| Command R+ (08-2024) | 48 | $2.50 | $10.00 |
| Command R (08-2024) | 48 | $0.150 | $0.600 |
| North Mini Code (free) | 40 | Free | Free |
| Command R7B (12-2024) | 35 | $0.037 | $0.150 |
| Model | Score | Input $/M | Output $/M |
|---|---|---|---|
| DeepSeek V4 Pro | 86 | $0.435 | $0.870 |
| DeepSeek V3.2 | 81 | $0.229 | $0.343 |
| R1 0528 | 79 | $0.500 | $2.15 |
| DeepSeek V4 Flash | 77 | $0.090 | $0.180 |
| R1 | 74 | $0.700 | $2.50 |
| DeepSeek V3 0324 | 71 | $0.200 | $0.770 |
| DeepSeek V3.2 Exp | 70 | $0.270 | $0.410 |
| DeepSeek V3.1 | 69 | $0.210 | $0.790 |
| DeepSeek V3.1 Terminus | 69 | $0.270 | $0.950 |
| DeepSeek V3 | 69 | $0.200 | $0.800 |
| R1 Distill Llama 70B | 41 | $0.800 | $0.800 |
Compare any two AI providers side-by-side.