DeepSeek: DeepSeek V4 Pro 0813

deepseek/deepseek-v4-pro-0813

DeepSeek V4 Pro 0813 is an open-weight model from DeepSeek, ranking #74 of 248 on the TryAii Score. It's strongest on knowledge (#19 on GPQA). It's served by 16 independent hosts from $0.264/$0.792 per million input/output tokens — about a third of what models at its level usually charge.

Prompt $/1M

$0.264

Completion $/1M

$0.792

Context

1,048,576 tok

Best provider

89 tok/s · CoreWeave

Specs

Model id: deepseek/deepseek-v4-pro-0813
Display name: DeepSeek: DeepSeek V4 Pro 0813
Modality: text->text
Tokenizer: DeepSeek
Instruct type: —
Context length: 1,048,576
Max completion: 384,000
Prompt $/1M: $0.264
Completion $/1M: $0.792
Free tier: no

Capabilities

Function callingno
JSON modeno
System promptno
Visionno
Moderatedno

Benchmark scores

Each axis is one benchmark, normalized to the best model in the field.

No benchmark data for this model.

Speed by provider

Provider	tok/s	TTFT	Total latency
CoreWeavefastest	89	1,038 ms	1,435 ms
Sail Research	88	1,373 ms	2,260 ms
Together	80	2,842 ms	3,029 ms
Cloudflare	62	1,544 ms	2,244 ms
Fireworks	62	1,267 ms	1,884 ms
Alibaba	60	1,324 ms	1,613 ms
DeepInfra	59	840 ms	1,040 ms
SiliconFlow	56	1,280 ms	1,580 ms
NextBit	55	3,283 ms	3,855 ms
StreamLake	54	2,744 ms	3,430 ms
Parasail	53	1,020 ms	1,555 ms
Phala	52	1,505 ms	1,915 ms
Novita	44	1,511 ms	1,752 ms
BaseTen	38	520 ms	625 ms
GMICloud	38	7,572 ms	8,926 ms
DigitalOcean	36	1,089 ms	1,430 ms
DeepSeek	20	1,152 ms	1,269 ms
deepseekfastest	—	—	—