DeepSeek

DeepSeek chat model for instruction following, coding, and analysis

Context

1,000,000

Output

384,000

Offers

8

High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes

Context

8,192

Output

8,192

Offers

0

Classic open reasoning model for transparent math, coding, and deliberate problem solving

Context

128,000

Output

32,768

Offers

18

R1 reasoning distilled into Qwen 2.5 32B for efficient open-weight step-by-step problem solving

Context

131,072

Output

32,768

Offers

0

DeepSeek reasoning model for multi-step analysis, math, coding, and tools

Context

1,000,000

Output

384,000

Offers

5

Open DeepSeek MoE chat model for coding, math, and general reasoning

Context

131,072

Output

8,192

Offers

0

March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding

Context

163,840

Output

163,840

Offers

0

Hybrid-reasoning DeepSeek model with thinking and non-thinking modes

Context

131,072

Output

8,192

Offers

0

Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use

Context

128,000

Output

64,000

Offers

0

Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work

Context

1,000,000

Output

384,000

Offers

40

Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding

Context

1,000,000

Output

384,000

Offers

0

Open MoE flagship with million-token context for coding and long agent runs

Context

1,000,000

Output

384,000

Offers

47

DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes

Context

1,000,000

Output

384,000

Offers

0