Understanding Pricing
Learn how pricing works on Glyphrun.
Pricing Models
Glyphs can use different pricing models:
Per-Run Pricing
Fixed cost per execution:
- Example: $0.05 per run
- Predictable - Know the cost upfront
- Most common - Used by most glyphs
Per-Token Pricing
Cost based on input/output tokens:
- Example: $0.001 per 1K tokens
- Variable - Depends on content length
- Used by: Some prompt glyphs
Per-Second Pricing
Cost based on execution time:
- Example: $0.01 per second
- Variable - Depends on task complexity
- Used by: Some agent glyphs
Free
No cost to run:
- Good for: Trying new glyphs
- May have: Usage limits
- Creators use for: Building reputation
Model Tiers (Prompts)
Prompt glyphs offer different model tiers:
| Tier | Price | Models | Best For |
|---|---|---|---|
| Fast | $0.02/run | GPT-4o Mini, Claude Haiku, Llama 8B | Simple tasks |
| Balanced | $0.05/run | GPT-4o, Claude 3.5 Sonnet, Llama 70B | Most use cases |
| Reasoning | $0.10/run | o1-preview, Claude Opus, Llama 405B | Complex analysis |
Cost Estimation
Before Running
The glyph page shows:
- Base price - Minimum cost
- Model tier prices - For prompts
- Estimated cost - Based on typical usage
During Execution
The Run button shows:
- Exact cost - What will be charged
- Your balance - Current available funds
What Affects Cost
For Prompts
- Model tier - Fast < Balanced < Reasoning
- Input length - More text = higher cost
- Output length - Longer responses cost more
For Agents
- Base price - Set by creator
- Execution time - For time-based pricing
- External API costs - Built into price
For Tools
- Base price - Set by creator
- API rate limits - May affect pricing
- Data volume - For data-heavy tools
Payment Flow
How It Works
- Lock - Amount is locked when you click Run
- Execute - Glyph runs
- Success - Locked amount is deducted
- Failure - Locked amount is released back
Failed Runs
If a glyph fails:
- You’re NOT charged
- Locked funds are released
- Balance is restored immediately
Creator Earnings
When you run a glyph:
- Creator receives - Majority of the payment
- Platform fee - Small percentage
- Instant settlement - Creators paid immediately
Tips for Cost Optimization
Choose the Right Tier
- Use Fast for simple tasks
- Use Balanced for most work
- Save Reasoning for complex analysis
Test with Examples
- Check Examples tab first
- Understand expected costs
- Start with simpler inputs
Monitor Spending
- Set spending limits
- Review transaction history
- Track monthly spending
Comparing Costs
Traditional AI APIs
| Service | Cost per 1K tokens |
|---|---|
| OpenAI GPT-4 | ~$0.03 |
| Anthropic Claude | ~$0.015 |
| Direct API | Variable |
Glyphrun
- Bundled pricing - No token counting
- Predictable costs - Fixed per-run prices
- No API keys - Use immediately
Next Steps
Last updated on: