Suivez les actualités des modèles DeepSeek, de l’API, des tarifs et des outils pour développeurs, ainsi que les nouvelles de l’IA. DeepSeek V4 Hub est une publication indépendante.
Checked on July 8, 2026: DeepSeek's official docs sitemap currently lists dedicated thinking-mode sample pages for both non-streaming and streaming chat completions, clarifying when `reasoning_content` is just captured, when it is ignored in ordinary follow-up turns, and how `delta.reasoning_content` should be accumulated separately from visible answer text.
Lire l’articleChecked on July 6, 2026: DeepSeek's official docs currently expose both a Python thinking-mode tool-call sample and a separate output transcript page, giving developers a first-party reference for tool schemas, chained calls, blank content turns, and reasoning_content replay behavior.
Lire l’articleChecked on July 5, 2026: DeepSeek's official homepage still anchors the official `@deepseek_ai` X presence, but the strongest current first-party developer signal is in the API docs and changelog, which both keep the July 24, 2026 retirement warning for `deepseek-chat` and `deepseek-reasoner` while pointing users to explicit `deepseek-v4-flash` and `deepseek-v4-pro` model names.
Lire l’articleChecked on July 2, 2026: OpenAI's Codex docs now document custom model providers through `model_provider`, `model_providers`, `base_url`, and `env_key`. That means DeepSeek can be framed as a third-party Codex route: use our DeepSeek access, point Codex at the configured DeepSeek provider, and keep the known Responses API, profile, Desktop picker, and V4 thinking-mode checks in place.
Lire l’articleChecked on July 2, 2026: DeepSeek's official API reference documents a `GET /models` endpoint that currently examples only `deepseek-v4-flash` and `deepseek-v4-pro`, plus a `GET /user/balance` endpoint with `is_available`, currency, granted balance, and topped-up balance fields for direct budget checks.
Lire l’articleChecked on July 1, 2026: DeepSeek's official X account published a permanent-discount update, and the current official Models & Pricing page still shows the lower DeepSeek V4 Flash and Pro token rates together with the existing 1M-context and 384K max-output contract.
Lire l’articleChecked on June 30, 2026: DeepSeek's official Token & Token Usage page says tokens are the billing unit, gives rough heuristics of about 0.3 token per English character and 0.6 token per Chinese character, and links a `deepseek_tokenizer.zip` package for offline token estimation while still treating API `usage` output as the real source of truth.
Lire l’articleChecked on June 30, 2026: Chinese financial and technology outlets report that DeepSeek upgrade emails point to a mid-July V4 formal release and a peak-valley API pricing policy. Current public DeepSeek docs still show the flat V4 Pro and Flash price table, so this item treats the peak-hour policy as a reported pricing update rather than a storefront or inventory change.
Lire l’articleChecked on June 29, 2026: DeepSeek's official Tool Calls guide documents OpenAI-compatible function calling, a Beta-only `strict` mode for exact schema matching, concrete supported and unsupported JSON Schema rules, and a required `tool` role replay pattern after execution.
Lire l’articleChecked on June 28, 2026: DeepSeek's official Models & Pricing and Rate Limit pages currently show a 1M context window, 384K max output, per-1M token pricing, and account-level concurrency limits of 2500 for DeepSeek V4 Flash and 500 for DeepSeek V4 Pro, with 429s when that cap is exceeded.
Lire l’articleChecked on June 27, 2026: DeepSeek's official multi-round conversation guide says `/chat/completions` does not store context server-side, so each follow-up call must resend the prior conversation history, including earlier assistant output, before the next user turn.
Lire l’articleChecked on June 26, 2026: DeepSeek's official WorkBuddy or CodeBuddy page now documents exact user-level and project-level `models.json` paths, `https://api.deepseek.com/v1/chat/completions` as the endpoint, a UTF-8 without BOM requirement, and a restart-first troubleshooting flow for V4 Pro and V4 Flash.
Lire l’article