chat-deepseek
GitHub调用 ZeroGPU API 使用 deepseek-v4-flash-0731 模型，适用于代码读写、重构及多步自动化规划。提供执行命令、隐私警告及成本提示，输出原始回复。
触发场景
安装
npx skills add zerogpu/zerogpu-router --skill chat-deepseek -g -y
SKILL.md
Frontmatter
{
"name": "chat-deepseek",
"metadata": {
"openclaw": {
"install": [
{
"bins": [
"zerogpu"
],
"kind": "node",
"package": "zerogpu-cli"
}
],
"requires": {
"bins": [
"zerogpu"
]
}
}
},
"description": "Chat with deepseek-v4-flash-0731, a 284B MoE model (13B active per token) with a 1M-token context window, tuned for coding and agentic workflows. Use for reading or writing code across a large codebase, porting and refactoring, or planning multi-step automation. Cheaper than chat-glm, with four times its context.",
"allowed-tools": "Bash(zerogpu chat_completions *)",
"argument-hint": "<text> [-i <instructions>]"
}
Sends your input to ZeroGPU's hosted API for inference — this is not local processing. Don't pass secrets, credentials, or regulated data you aren't cleared to share with a third party. See the plugin README's "Data & privacy" section.
Call deepseek-v4-flash-0731. Run this with the exec tool, pasting the user's prompt into the heredoc verbatim — no escaping or quoting required (the quoted heredoc handles every shell metacharacter, newline, quote, and paren safely):
zerogpu chat_completions -m deepseek-v4-flash-0731 <<'ZGPU_END_OF_INPUT'
<the user's prompt, verbatim>
ZGPU_END_OF_INPUT
If the user supplied system instructions, add -i '<instructions>' after the model flag, writing any ' inside them as '\''.
At $0.16 / $0.38 per 1M input/output tokens this is the cheaper of the two 1M-context models on the platform — roughly a seventh of chat-glm on input and a ninth on output. Prefer it whenever the task is code or tool-use rather than sheer input size. For a prompt that fits in 131K tokens, chat ($0.15 / $0.60) matches it on input and costs more on output.
Output is the assistant's answer as plain text — the model's reasoning trace comes back in a separate field and is not printed. Relay the answer as-is — do not rewrite or expand it.
Savings note: only if the command output literally contains a line starting with 💰 ZeroGPU savings, append that exact line, unchanged, as the last line of your reply. If no such line is present, say nothing about savings and do not mention or suggest the cost-savings skill — this note is intentionally occasional, not shown every time.
版本历史
-
4e1b070
当前 2026-09-22 08:07
从 deepseek-v4-flash 升级至 deepseek-v4-flash-0731，调整定价策略，并更新命令调用方式以适配新版 CLI。
- cee9321 2026-08-02 21:15
