Hi Pinecone Team,
We’re currently developing a chatbot for a customer and plan to integrate Pinecone Assistant MCP server along with our own GPT-4o-mini-powered agent (for final answers). We’ve completed a prototype using Pinecone Assistant and now want to move towards actual deployment.
To estimate costs accurately, we’d appreciate help based on the following usage assumptions:
| Metric | Value |
|---|---|
| Total Users | 50,000 |
| Chats per User per Day | 10 |
| Context Tokens per Chat | 4,000 |
| Duration (in Days) | 10 |
We’ll only be using the following features from Pinecone Assistant:
- Storage
- Context Processed Tokens (via Context API)
- Hourly Rate (assistant hosting cost)
From the pricing chart, we understand:
- Context Processed Tokens: $5 per 1M tokens (Standard/Enterprise)
- Hourly Rate: $0.05 per Assistant per hour
- Storage: $3 per GB/month beyond included 1GB (Standard/Enterprise)
Could someone from the team help calculate a rough cost estimate for this usage? Specifically the total cost of context tokens and regarding the plan ideal for us ( between Standard and Enterprise ) .
When we did a quick math to get the cost of context API ( according $5 per 1M tokens ), the pricing came to be really high ( ~$100k ).
( ( 50000 * 10 * 4000 * 10) / 1000000 ) * 5 = 100000 is how we got the $100k cost.
Thanks in advance for your guidance!
Best regards,
Gaurav Pant