Chinese models in Claude Code and Codex

DeepSeek · Coding and reasoning models

Run Codex on DeepSeek V4

The one model on this list you can use with Codex today. Flash runs $0.14 in / $0.28 out per million tokens — five to fifteen times less than OpenAI for the same work — and both tiers carry a 1M-token context.

Works today

Yes. We ship a small macOS and Windows app that points Codex CLI, Codex Desktop and the editor plugins at api.deepseek.com using your own DeepSeek key. The key stays in your system keychain and nothing routes through us.

See the app

The numbers

Built by
DeepSeek
Released
Jan 20, 2026
Context window
1,000,000 tokens
Per 1M tokens (in / out)
$0.14 / $0.28
In HaryAI
Works today

List price from the vendor's own API — cache-miss input and output. Third-party hosts are often cheaper, and cached reads change the real number.

What it takes to use it

Codex reads ~/.codex/config.toml, so whatever changes that file changes every Codex surface at once. You can edit it by hand — set the provider base URL to api.deepseek.com, drop in your key — or use our app, which writes the same file, keeps the key in the keychain and puts token spend in the status bar. Tool calling, file edits and shell exec behave the way they do on OpenAI.

Figures checked 2026-08-06 against vendor pricing pages and launch coverage. Models re-price often — if something here is stale, tell us.

Run Codex on DeepSeek V4 · HaryAI