What You Need to Know
Poppy gives you access to the best AI models - each one built for different jobs. Think of them like different tools in a toolbox. You wouldn't use a hammer to cut wood, right? Same deal here.
The Models at a Glance
Model | Best Use Cases | Credit Spending | Token Limit |
|
Claude 4 Sonnet | Best all-rounder for creators. Handles TikTok captions, Reel hooks, newsletters and video scripts with great speed and quality. | Normal | 200k |
|
Claude 4.7 Opus | Most powerful Claude model. For the heaviest creative work including book writing, full course creation and brand strategy that needs deep nuanced thinking. | Normal | 200k |
|
GPT-5 | Most advanced reasoning, creativity, multimodal, better memory & adaptability | Normal | 400k |
|
GPT-4o | Fast multimodal output for live content workflows, image-driven posts and quick social media turnaround. | Normal | 128k |
|
Gemini 2.5 Pro | Best for processing large source material. Full video transcripts, research docs and massive datasets to build content series from. | Normal | 1 million |
|
Grok 3 | Specialized in snappy, culturally aware content. Great for social media copy, trending topic scripts and punchy short-form writing. | Normal | ~128k |
|
GPT-5 Mini | Budget-friendly for daily posting. Fast reliable quality for routine captions, short scripts and simple creative tasks. | Low | ~400k |
|
Image Generation Models
Model | Best Use Cases | Credit Spending |
|
GPT Image 1 | Quick thumbnails, simple icons and rough visual concepts for ideation or placeholder content. | Best Price |
|
NanoBanano | Social media graphics, product mockups and single-subject illustrations. Solid quality for everyday posts. | Cheap |
|
NanoBanano Pro | High-detail image generation with complex scenes, strong style control and brand-consistent visuals for content that needs to stand out. | Normal |
|
What Are Token Limits?
Think of tokens as the "memory capacity" of each AI model.
Here's what counts toward your token limit:
Everything you have attached to your chat (documents, videos, links, etc.)
Your chat history with that AI
So if you've been in one chat for a while and your conversation history is super long, you might hit the token limit even with nothing attached to your chat.
When You Hit the Overload Limit
If you see an "overload limit" error, don't panic. It just means you're trying to feed the AI more content than it can process at once.
For complete troubleshooting steps, check out our Overload Limit Troubleshooting Guide
How to Switch AI Models
Switching between models is dead simple. In your chat, look for the dropdown menu right under the text bar - it shows which AI model you're currently using.
To test different models:
Send your prompt with the current AI model
Click the dropdown and select a different model
Send the same prompt again
Compare the results to see which model works better for your specific task
This makes it super easy to find the perfect AI for whatever you're working on.