DeepSeek V4 Pro 0813
by CoreWeave · family deepseek-thinking · listed Aug 2026RTW
DeepSeek V4-Pro-0813 is a 1.6T-parameter MoE model excelling at advanced reasoning, coding, and complex agentic workloads.
Current rates USD per 1M tokens
Input$1.31
Output$3.96
Cache read$0.044
Cache write—
Against the market
On the record
- Model id
deepseek-ai/DeepSeek-V4-Pro-0813- Context window
- 1.0M tokens
- Max output
- 1.0M tokens
- Input modalities
- text
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2026-08-13
- Last updated
- 2026-08-22
- Provider docs
- docs.wandb.ai/inference
Back of the envelope
≈
Also listed at
- SCNet Token Planfree in · free out
- Alibaba Token Planfree in · free out
- Alibaba Token Plan (China)free in · free out
- OpenRouter$0.115 in · $4.00 out
- NanoGPT$0.284 in · $0.851 out
- RunInfra$0.60 in · $1.90 out
- Eden AI$0.66 in · $1.98 out
- OrcaRouter$0.66 in · $1.98 out
- Vercel AI Gateway$0.66 in · $1.98 out
- AIHubMix$0.692 in · $2.08 out
- IteraCompute$0.79 in · $2.38 out
- Boundless$1.04 in · $2.08 out
- Standard Compute$1.12 in · $3.37 out
- Deep Infra$1.30 in · $2.60 out
- Nebius Token Factory$1.32 in · $3.96 out