DeepSeek: DeepSeek V3.2
deepseek/deepseek-v3.2
DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
Modalities
In / Out Price
$0.259 / $0.42 per 1M
Context
164K
Released
Dec 1, 2025
Knowledge Cutoff
—
Providers
Different carriers host the same model. Reroute routes each request to the best one on price and uptime, and fails over to the next if a carrier goes down.
| Provider | Context | Input /M | Output /M | Cache read /M | Latency | Throughput | Uptime |
|---|---|---|---|---|---|---|---|
 GMICloudfp8 | 164K | $0.209 | $0.31 | $0.0216 | — | — | 100.00% |
 SiliconFlowfp8 | 164K | $0.259 | $0.42 | $0.135 | — | — | 93.33% |
 DeepInfrafp4 | 164K | $0.26 | $0.38 | $0.13 | — | — | 100.00% |
 AtlasCloudfp8 | 164K | $0.26 | $0.38 | $0.13 | — | — | 73.33% |
 Venice | 160K | $0.268 | $0.39 | $0.13 | — | — | 100.00% |
 Baidufp8 | 131K | $0.28 | $0.42 | $0.028 | — | — | 100.00% |
 DigitalOcean | 164K | $0.30 | $0.96 | $0.09 | — | — | 100.00% |
 Alibabafp8 | 131K | $0.37 | $1.11 | $0.0741 | — | — | 93.33% |
 Friendli | 164K | $0.50 | $1.50 | $0.25 | — | — | 100.00% |
 Google | 164K | $0.56 | $1.68 | — | — | — | 100.00% |
 Phala | 164K | $1 | $1 | $0.50 | — | — | 100.00% |
 Mara | 33K | $3 | $4.50 | — | — | — | 100.00% |
 SambaNova | 33K | $3 | $4.50 | — | — | — | — |