LLMs-Universal-Life-Science-and-Clinical-Skills- deepseek-api-operations-2026

Integrate and operate DeepSeek APIs with current docs and compatibility guidance. Use when implementing DeepSeek chat, reasoning, tool calling, or FIM workflows through its OpenAI-compatible API.

install

source · Clone the upstream repo

git clone https://github.com/mdbabumiamssm/LLMs-Universal-Life-Science-and-Clinical-Skills-

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/mdbabumiamssm/LLMs-Universal-Life-Science-and-Clinical-Skills- "$T" && mkdir -p ~/.claude/skills && cp -r "$T/Skills/AI_Providers/DeepSeek_API_Operations_2026" ~/.claude/skills/mdbabumiamssm-llms-universal-life-science-and-clinical-skills-deepseek-api-opera && rm -rf "$T"

manifest: Skills/AI_Providers/DeepSeek_API_Operations_2026/SKILL.md

source content

DeepSeek API Operations (2026)

Workflow

Start from DeepSeek's OpenAI-compatible API assumptions and document the base URL explicitly.
Choose between
```
deepseek-chat
```
,
```
deepseek-reasoner
```
, or beta/FIM paths based on the workload.
Validate tool calling, JSON mode, and streaming behavior against the official docs before using the OpenAI SDK as a drop-in.
Plan around long-lived requests and server-side timing behavior rather than assuming OpenAI-style latency.
Run a smoke test that proves model output, reasoning output, and tool output behave as expected.

Output Requirements

State the chosen model and base URL.
State the compatibility assumptions.
State one timeout or retry guardrail.