DeepSeek prompt generator
Prompts for DeepSeek’s chat and reasoning models. Reasoning models think before they answer, so give them a clear goal and the facts they need, and skip step-by-step scaffolding.
Fill in a few fields and your prompt is set here, line by line.
Prompt score
Underspecified
0 words
- Clear task verb missingStart with an explicit action: “Write…”, “Summarize…”, “Compare…”.
- Background or context missingSay who it is for and why you need it; models fill gaps with generic guesses.
- Role or perspective missingGive a perspective: “You are a senior UX writer…”. It sets vocabulary and depth.
- Output format stated missingName the shape of the answer: table, bullets, JSON, 3 paragraphs.
- Length or scope limit missingCap the length: “under 150 words”, “5 bullets”.
- Constraints or rules missingAdd rules: what to avoid, what must be included, reading level.
- Example or reference missingShow one example of the output you want. Few-shot beats description.
- Specific, not vague missingReplace vague words (good, nice, interesting) with concrete criteria.
Let AI polish it
Sends your filled-in fields to the AI helper, which writes a complete prompt and marks gaps as [placeholders].
Prompting DeepSeek well
Keep reasoning-model prompts direct
DeepSeek’s reasoning models work through problems on their own. State the goal, the facts and the expected output format; long “think step by step” instructions add little.
Put everything in the user message
DeepSeek’s guidance for its R1 reasoning model recommended placing instructions in the user prompt rather than a system prompt. The generator’s single-message output suits this.
Ask for a checkable final answer
For maths or code, ask for the final answer in a fixed format (e.g. “Final answer: …”) so it is easy to verify.
New to structured prompts? The prompt engineering guide explains each part, and the prompt optimizer scores any prompt you already have.
Questions
How should I prompt DeepSeek R1 or other reasoning models?
Be direct: describe the problem, give the necessary context and say how you want the answer formatted. Avoid few-shot examples and heavy step-by-step instructions, which DeepSeek reported can reduce reasoning-model performance.
Does a DeepSeek prompt differ from a ChatGPT prompt?
For chat models, barely. For reasoning models, keep prompts shorter and more goal-focused, and let the model plan its own steps.
Is this tool affiliated with DeepSeek?
No. CraftPrompts is an independent, free prompt studio that works with any model.
Prompt generators for other models
ChatGPT · Claude · Gemini · Grok · Perplexity · Nano Banana · Veo 3 · Suno · Midjourney · Stable Diffusion
More tools in the studio
- GeneratorPrompt GeneratorSet a text prompt from role, task, context, format and constraints.
- OptimizerPrompt OptimizerScore any prompt on 8 checks, then rewrite it and compare.
- ChatGPTChatGPT Prompt GeneratorMarkdown-sectioned prompts tuned for ChatGPT.
- ClaudeClaude Prompt GeneratorXML-tagged prompts the way Anthropic recommends.
- SystemSystem Prompt GeneratorInstructions for custom GPTs, assistants and agents.
- Image → promptImage to PromptUpload a picture, get the prompt that recreates its look.
- ImageImage Prompt GeneratorNatural-language prompts for Flux, GPT Image, DALL·E and Imagen.
- MidjourneyMidjourney Prompt BuilderDescriptors plus --ar, --stylize, --chaos and more.
- Stable DiffusionStable Diffusion Prompt GeneratorWeighted tags, negative prompt, CFG and steps.
- VideoVideo Prompt GeneratorShot-by-shot prompts for Sora, Veo, Runway and Kling.
- TokensToken CounterExact GPT token counts, context fit and cost.
- GeminiGemini Prompt GeneratorPersona, task, context and format for Google Gemini.