DeepSeek

Summary pending.

• DeepSeek V4 Flash model now available with 13B active parameters optimized for coding, reasoning, and agent workflows
• Context window supports up to 1,048,576 tokens
• Pricing set at $0.09 per million input tokens and $0.18 per million output tokens

• DeepSeek V4 Pro model is now available with 1.6T total parameters and 49B activated parameters
• Supports 1M-token context window for processing longer documents and conversations
• Pricing set at $0.43 per million input tokens and $0.87 per million output tokens

• DeepSeek V4 Flash is a new efficiency-optimized Mixture-of-Experts model with 284B total parameters and 13B activated parameters
• Supports a 1M-token context window for processing large documents
• Priced at $0.14 per million input tokens and $0.28 per million output tokens

• DeepSeek-V3.2 introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism for improved efficiency
• Context window expanded to 163,840 tokens
• Pricing set at $0.27 per million input tokens and $0.40 per million output tokens

• DeepSeek-V3.2-Exp introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism
• Context window expanded to 163,840 tokens
• Knowledge cutoff updated to 2025-07-31
• Pricing set at $0.27 per million input tokens and $0.41 per million output tokens

• DeepSeek V3.1 Terminus addresses reported issues with language consistency and agent capabilities
• Context window remains at 163,840 tokens
• Knowledge cutoff updated to 2025-03-31
• Pricing adjusted to $0.27 per million tokens input / $1.00 per million tokens output

• DeepSeek-V3.1 is now available as a hybrid reasoning model supporting both thinking and non-thinking modes
• Context window expanded to 163,840 tokens
• Knowledge cutoff updated to 2025-03-31
• Pricing set at $0.25 per million input tokens and $0.95 per million output tokens

• DeepSeek R1 updated to May 28th version with performance comparable to OpenAI o1
• Model now has fully open reasoning tokens available
• Context window is 163,840 tokens with knowledge cutoff of March 31, 2025
• Pricing set at $0.50 per million input tokens and $2.15 per million output tokens

• DeepSeek V3 model released with 685B parameters and improved performance
• Context window expanded to 163,840 tokens
• Knowledge cutoff updated to 2024-07-31
• Pricing set at $0.27 per million input tokens and $1.12 per million output tokens

• DeepSeek R1 Distill Llama 70B is now available as a distilled model based on Llama-3.3-70B-Instruct using DeepSeek R1 outputs
• Model has an 8,192 token context window with knowledge cutoff at 2024-07-31
• Pricing is $0.80 per million tokens for both input and output

• DeepSeek R1 model released with performance comparable to OpenAI o1
• Fully open-sourced with visible reasoning tokens
• 671B parameter model with 163,840 token context window
• Knowledge cutoff at July 31, 2024

• DeepSeek-V3 model is now available with improved instruction following and coding abilities
• Context window expanded to 163,840 tokens
• Knowledge cutoff updated to 2024-07-31
• Pricing set at $0.26 per million input tokens and $1.03 per million output tokens