DeepSeek V3

#7 · Chat

Open-weight, extremely cheap reasoning model from DeepSeek with surprising quality.

Last updated: 2026-08-14

Editorial note

The open-weight value king. Quality that rivals far pricier models at a fraction of the cost.

DeepSeek V3 vs…

DeepSeek V3 vs GPT-5

OpenAI

OpenAI

▲ 12

Flagship multimodal reasoning model from OpenAI with 200K context and state-of-the-art tool use.

Anthropic

▲ 8

Anthropic’s most capable model with a 1M context window, best for long-form analysis and code review.

Google

▲ 5

Google’s flagship with a 2M context window, native multimodal input and cheap inference.