Loading...
Loading...
DeepSeek V3 — 685B MoE model achieving frontier-tier coding and reasoning at a fraction of closed-source API costs.
$0.135
per 1M input tokens
—
top score across benchmarks
10
APIs serving this model
128K
max token window
Looking for alternatives?
Find cheaper or faster models in the same class.
See DeepSeek V3 alternatives →deepseek DeepSeek: DeepSeek V3.2 Speciale | $0.287 | $0.431 | — | 164k |
deepseek-ai Deepseek V3 0324 | $1.250 | $1.250 | — | 164k |
deepseek DeepSeek: DeepSeek V3 0324 | — | — | 18ms | 164k |
deepseek DeepSeek: DeepSeek V3 | — | — | 16ms | 131k |
deepseek DeepSeek: DeepSeek V3.1 Terminus | — | — | 53ms | 164k |
deepseek DeepSeek: DeepSeek V3.2 | — | — | 33ms | 131k |
deepseek DeepSeek: DeepSeek V3.2 Exp | — | — | 74ms | 164k |
deepseek DeepSeek: DeepSeek V3.1 | — | — | 63ms | 164k |
* Free tier or self-hosted option available. Compute costs apply for self-hosting.