DEV Community

#deepseek

DeepSeek is a Chinese artificial intelligence company that has developed an open-source AI model, DeepSeek-R1, which rivals leading models like OpenAI's o1 in performance, yet was created in under two months at a cost of less than $6 million, significantly lower than its competitors.

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
Run DeepSeek V4 on Your Own Hardware With DwarfStar, the Redis Creator's New Inference Engine

Run DeepSeek V4 on Your Own Hardware With DwarfStar, the Redis Creator's New Inference Engine

1
Comments
7 min read
A new agent, and nothing else moved

A new agent, and nothing else moved

Comments
4 min read
Your DeepSeek V4 Bill Can Swing 2x by the Hour — the Actual Math Across Four APIs

Your DeepSeek V4 Bill Can Swing 2x by the Hour — the Actual Math Across Four APIs

Comments
4 min read
LLM vs Agent vs Harness, Explained by a Caveman

LLM vs Agent vs Harness, Explained by a Caveman

Comments
2 min read
Codex + DeepSeek: why "No tool output found for tool call" kills your session (and how to avoid it)

Codex + DeepSeek: why "No tool output found for tool call" kills your session (and how to avoid it)

2
Comments
7 min read
SpellBook of Skill: The Twin Project Nobody Asked For, But I Built Anyway 🧙ðŸŧ‍♂ïļ

SpellBook of Skill: The Twin Project Nobody Asked For, But I Built Anyway 🧙ðŸŧ‍♂ïļ

5
Comments
4 min read
codsh 0.23.0 — Mission Contract: `/ship` that won't silently drift

codsh 0.23.0 — Mission Contract: `/ship` that won't silently drift

Comments
3 min read
DeepSeek V4.1 Flash API Cost: Matches V4 Pro at 3-6x Less per Answer

DeepSeek V4.1 Flash API Cost: Matches V4 Pro at 3-6x Less per Answer

Comments
12 min read
āļĢāļēāļ„āļē cache hit 0.003 āļ”āļ­āļĨāļĨāļēāļĢāđŒ āļ—āļĩāđˆāđ€āļ›āļĨāļĩāđˆāļĒāļ™āļ§āļīāļ˜āļĩāđ€āļ—āļĩāļĒāļšāđ‚āļĄāđ€āļ”āļĨāļ—āļąāđ‰āļ‡āļ•āļĨāļēāļ”

āļĢāļēāļ„āļē cache hit 0.003 āļ”āļ­āļĨāļĨāļēāļĢāđŒ āļ—āļĩāđˆāđ€āļ›āļĨāļĩāđˆāļĒāļ™āļ§āļīāļ˜āļĩāđ€āļ—āļĩāļĒāļšāđ‚āļĄāđ€āļ”āļĨāļ—āļąāđ‰āļ‡āļ•āļĨāļēāļ”

Comments
2 min read
āļ—āļģāđ„āļĄ KV cache āļ–āļķāļ‡āļāļĨāļēāļĒāđ€āļ›āđ‡āļ™āļŦāļąāļ§āđƒāļˆāđ€āļĻāļĢāļĐāļāļĻāļēāļŠāļ•āļĢāđŒāļ‚āļ­āļ‡ AI agent āđƒāļ™āļ›āļĩ 2026

āļ—āļģāđ„āļĄ KV cache āļ–āļķāļ‡āļāļĨāļēāļĒāđ€āļ›āđ‡āļ™āļŦāļąāļ§āđƒāļˆāđ€āļĻāļĢāļĐāļāļĻāļēāļŠāļ•āļĢāđŒāļ‚āļ­āļ‡ AI agent āđƒāļ™āļ›āļĩ 2026

Comments
2 min read
DeepSeek R1: The Open-Source Reasoning Revolution That Changes Everything

DeepSeek R1: The Open-Source Reasoning Revolution That Changes Everything

Comments
2 min read
DeepSeek as the default model on ChatGPT Codex

DeepSeek as the default model on ChatGPT Codex

1
Comments
4 min read
Building Session-Level Observability for DeepSeek Harness with a Community Plugin

Building Session-Level Observability for DeepSeek Harness with a Community Plugin

1
Comments
5 min read
DeepSeek plans to order 160,000 Huawei chips for new center

DeepSeek plans to order 160,000 Huawei chips for new center

5
Comments
3 min read
codsh 0.22.1 — /ship grill cards wrap long questions (DeepSeek terminal agent)

codsh 0.22.1 — /ship grill cards wrap long questions (DeepSeek terminal agent)

Comments
1 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.