Provider comparison · Updated 10 Oct 2026
Cohere vs DeepSeek
6 models vs 13 models
DeepSeek is cheaper at the median: $0.27 against $0.30 per 1M input tokens across paid models, about 1.1× less. Its cheapest paid model, DeepSeek V4 Flash 0731 ($0.08), compares with Command R7B (12-2024) ($0.04) at Cohere. Cohere lists 6 models; DeepSeek lists 13 models.
Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
Cite
Cite this page
ModelsAtlas. “Cohere vs DeepSeek”. Retrieved 10 Oct 2026. https://modelsatlas.com/compare-providers/cohere/deepseek
At a glance
Standard tier, paid models · highlighted = better
Cohere
DeepSeek
Models listed
6
13
Cheapest input / 1M
$0.04
$0.08
Median in / out
$0.30 / $1.50
$0.27 / $0.95
Highest input / 1M
$2.50
$0.70
Largest context
256K
1.05M
Batch-tier models
0 of 6
1 of 13
Open-weight models
0
3
Free models
1
0
Vision models listed
0
1
Costs for real workloads
Cheapest paid model from each provider that fits the job · monthly
Derived from standard-tier prices, 30 days. “Fits the job” means the context window holds the prompt and reply. Excludes batch and caching.
Cheapest model for each capability
Input / output per 1M
| Capability | Cohere | DeepSeek |
|---|---|---|
| 1M context | None listed | DeepSeek V4 Flash 0731 $0.08 / $0.17 |
Line-up by price tier
Paid models · input per 1M
| Tier | Cohere | DeepSeek |
|---|---|---|
| Flagship $3+ | None listed | None listed |
| Mid $0.50–$3 | Command A $2.50 Command R+ (08-2024) $2.50 | R1 0528 $0.70 |
| Budget under $0.50 | Command R7B (12-2024) $0.04 Command R (08-2024) $0.15 Command A+ $0.30 | DeepSeek V4 Flash 0731 $0.08 DeepSeek V4 Flash Vision Exp $0.22 DeepSeek V3.1 $0.25 DeepSeek V3.2 $0.26 DeepSeek V3.1 Terminus $0.27 +3 more |
| Open weights any price | None listed | None listed |
Questions
Is Cohere or DeepSeek cheaper?
DeepSeek at the median: $0.27 against $0.30 for Cohere per 1M input tokens. Median is across each provider's paid standard-tier models, so a provider with many small models can look cheaper than its flagship is.
Which provider has more models?
DeepSeek lists more: Cohere 6, DeepSeek 13. Counts are models with a page in our catalogue.
Do both offer batch pricing?
Cohere: no batch tier listed. DeepSeek: 1 model list a batch tier. Where a model lists both, batch is a median 37% for DeepSeek.
Which has the larger context window?
DeepSeek has the larger: Command A (256K tokens) at Cohere, DeepSeek V4 Flash 0731 (1.05M) at DeepSeek.
Does either provider offer open-weight models?
Cohere: none marked open-weight. DeepSeek: 3 open-weight models.
Does either have free models?
Cohere: 1 model listed at no per-token charge. DeepSeek: none.
Browse more providers
Keep going in your AI
Opens with this page as the source. We add nothing else to the prompt.
- Is Cohere or DeepSeek cheaper?
- Which provider has more models?
- Do both offer batch pricing?