G2

Bob V.

Founder and CTO

Information Technology and Services

Small-Business (50 or fewer emp.)

6/26/2026

"Powerful Open-Source AI with 1M Token Context and Strong Reasoning"

4/5

What do you like best about deepseek-v4?

The standout feature is the 1 million token context window, which is made practical by the highly efficient hybrid attention architecture. DeepSeek-V4-Pro uses only 27% of single-token inference FLOPs and 10% of KV cache compared to its predecessor, making long-context tasks genuinely feasible. The open-source MIT license is also a major plus. Review collected by and hosted on G2.com.

What do you dislike about deepseek-v4?

The main downside is the very high hardware requirement to self-host — with 1.6T total parameters for the Pro model, running it locally demands significant GPU infrastructure. Additionally, the custom chat template (not standard Jinja format) adds some integration complexity for developers. Review collected by and hosted on G2.com.

What problems is deepseek-v4 solving and how is that benefiting you?

DeepSeek-V4 solves the challenge of processing very long documents and codebases in a single context, thanks to its 1 million token window. It also enables cost-effective advanced reasoning — the MoE architecture keeps compute costs low while delivering top-tier performance in coding, math, and agentic tasks. Review collected by and hosted on G2.com.