Skip to content
New this week26 prompts for Sora, Veo and Runway video workflows.
Explore video prompts→
AIPrompt123Prompt library

Search AIPrompt123

Search prompts, categories, platforms and repositories

Prompts tuned for the model you actually use

The same task needs different phrasing on different models. Claude prefers XML structure and explicit reasoning tags; Gemini needs citation instructions; Midjourney wants parameters instead of sentences. Each of the 25 hubs below collects the prompts that work on that platform, with notes on model quirks and a tuned copy of every template.

Chat & reasoning

7 platforms

ChatGPT

ChatGPT

The most widely used general assistant, with strong instruction following, browsing, file analysis and custom GPTs. Our ChatGPT prompts use explicit roles, output formats and short examples to keep answers predictable.

457 promptsOfficial site

Claude

Claude

Anthropic's assistant, strongest on long documents, careful reasoning and nuanced writing. Claude prompts work best with XML-style tags, a clear task statement and an invitation to think before answering.

462 promptsOfficial site

Gemini

Gemini

Google's multimodal assistant with grounding in Search, Workspace integrations and very long context. Gemini prompts benefit from source constraints, citation requests and explicit formatting.

71 promptsOfficial site

Grok

Grok

xAI's assistant with real-time access to X posts and a direct, irreverent tone. Grok prompts should specify tone explicitly, because its default voice is more casual than other models.

DeepSeek

DeepSeek

A low-cost reasoning-focused model family popular for maths, coding and structured analysis. DeepSeek prompts should state the reasoning budget and require a concise final answer.

19 promptsOfficial site

Microsoft's assistant across Windows, Edge and Microsoft 365. Copilot prompts should name the app context and the exact artefact you want, such as an Outlook reply or a PowerPoint outline.

13 promptsOfficial site

Llama

Llama

Meta's open-weight model family, available locally and through many providers. Llama prompts should be explicit and self-contained, because smaller variants have less instruction-following headroom.

Coding & builders

5 platforms

Cursor

Cursor

An AI-first code editor built on VS Code. Cursor prompts should reference files with @, state the acceptance test and constrain the diff size.

17 promptsOfficial site

GitHub Copilot

GH Copilot

GitHub's coding assistant for inline completion, chat and pull request review. GitHub Copilot prompts work best as short, context-rich comments and chat instructions.

Lovable

Lovable

A full-stack app builder that turns natural language into a working React app with a database. Lovable prompts need scope control, data model clarity and explicit acceptance criteria.

Bolt

Bolt

StackBlitz's in-browser app generator for instant prototypes. Bolt prompts should name the framework, keep to one feature at a time and specify the visual direction.

v0

v0

Vercel's generative UI tool for React and Tailwind interfaces. v0 prompts are design briefs: layout, components, states, responsive rules and accessibility.

Image generation

5 platforms

Midjourney

Midjourney

An image generator known for strong aesthetics and stylisation. Midjourney prompts follow a subject, style, lighting, composition formula and use parameters such as --ar and --style raw.

24 promptsOfficial site

DALL-E

DALL-E

OpenAI's image model inside ChatGPT and the API, strong at following complex scene descriptions and text rendering. DALL-E prompts read like a precise art brief.

16 promptsOfficial site

Stable Diffusion

Stable Diffusion

An open image model family with unrivalled control through LoRAs, ControlNet and inpainting. Prompts pair weighted keywords with negative prompts and technical settings.

Flux

Flux

Black Forest Labs' image family, excellent at photorealistic people, legible text and prompt adherence. Flux prompts favour natural language with camera and lighting detail.

15 promptsOfficial site

Leonardo

Leonardo

A game-art and asset-focused image platform with fine-tuned models, element LoRAs and canvas editing. Leonardo prompts name the fine-tuned model and art direction.

12 promptsOfficial site

Video generation

4 platforms

Sora

Sora

OpenAI's text-to-video model for cinematic short clips. Sora prompts read like a shot list: subject, action, camera, lighting, duration and continuity.

23 promptsOfficial site

Veo

Veo

Google DeepMind's video generator with strong physics, audio and prompt adherence. Veo prompts benefit from filmic detail and explicit sound direction.

19 promptsOfficial site

Runway

Runway

A video generation and editing suite with Gen models, motion brushes and frame control. Runway prompts pair a visual description with camera and motion instructions.

21 promptsOfficial site

Pika

Pika

A playful, fast video generator known for effects and social-first clips. Pika prompts are short, visual and effect-led.

12 promptsOfficial site

Music & audio

2 platforms

Suno

Suno

A text-to-music platform that generates full songs with vocals from a style prompt and lyrics. Suno prompts separate genre tags from structured lyric sections.

11 promptsOfficial site

ElevenLabs

ElevenLabs

A voice generation and cloning platform used for narration, audiobooks and agents. ElevenLabs prompts control delivery through text, punctuation, audio tags and voice settings.

15 promptsOfficial site

Research

2 platforms

Perplexity

Perplexity

An answer engine that searches the live web and cites sources. Perplexity prompts should demand source quality rules, recency windows and a clear separation between evidence and interpretation.

32 promptsOfficial site

NotebookLM

NotebookLM

Google's source-grounded research notebook with citations, summaries and audio overviews. NotebookLM prompts should constrain answers to the uploaded sources.

16 promptsOfficial site