DeepSeek 接入 Claude Code 已有官方路线，Codex 仍应写成自定义 provider 观察项

截至 2026 年 7 月 2 日，DeepSeek 官方 Claude Code 文档已经给出 Anthropic 端点、V4 Pro 主模型、Flash 子代理、Web Search 和 Claude 模型名映射。Codex 这边还没有对应的 DeepSeek 官方接入页，当前证据主要来自 openai/codex issue：thinking mode 的 reasoning_content、Desktop 模型选择、profile 路由和 auto-review 都仍有自定义 provider 摩擦。

中文摘要

阅读提示

这篇中文稿保留原始来源链接，并把 DeepSeek 官方发布、报道和市场传闻分开标注。购买相关判断仍以 /zh/pricing 的真实库存卡片为准；出现在新闻或基准中的模型不代表可购买。

英文原文

What changed after the June Claude Code item

Claude Code is no longer the messy part of the story. DeepSeek's own docs give teams a direct Anthropic-compatible route: point Claude Code at https://api.deepseek.com/anthropic, use deepseek-v4-pro[1m] for the main lane, keep deepseek-v4-flash for subagents, and set the effort level to max.

That is still worth a fresh update because the same DeepSeek page now carries three practical details that older community snippets often miss: Web Search can run through DeepSeek, Claude-style model names can map to V4 Pro or V4 Flash, and search-triggered summarization costs extra model tokens. This is a setup contract, not just a screenshot walkthrough.

Codex is a different story

Do not call Codex an official DeepSeek integration yet. No matching DeepSeek agent-integration page for OpenAI Codex showed up in the official docs checked for this update. The current signal is narrower and more operational: developers are trying DeepSeek through Codex custom providers, and the openai/codex issue tracker shows where that path still catches.

DeepSeek thinking mode can fail in multi-turn Codex sessions if reasoning_content is not passed back after tool-use turns.
Codex Desktop has separate reports around custom-provider model catalogs and picker visibility, especially when users expect GUI switching instead of manual config.toml edits.
A June report says Codex CLI 0.14.0 ignored --profile and custom model_providers in one setup, falling back to OpenAI instead of the configured DeepSeek-shaped provider.
Another report says Codex auto-review can send codex-auto-review to a custom DeepSeek provider, which DeepSeek rejects because it only recognizes its own V4 model names.

Those are useful developer signals, but they are not the same as DeepSeek saying 'Codex is supported' in its docs. The honest headline is: Claude Code has the official path; Codex is possible only through custom-provider work, and the rough edges are still visible.

How to use this on the site

Do not write another generic Claude Code setup page just to chase freshness. The existing Claude Code and Anthropic-routing pages already cover the official route. The useful addition is the boundary: when readers ask about Codex, point them to the OpenAI-format route, DeepSeek V4 model names, and the known custom-provider risks instead of promising a polished first-party flow.

This also keeps the commercial line clean. None of these sources creates a new stocked Coding Plan, a new provider card, or a plan migration. It is content and troubleshooting coverage only.