Cost transparency: what costs what
Every action in SERPmantics has a cost you know in advance. API v1 and MCP cost strictly the same, action by action. Creating guides and the EEAT analysis cost the same on every channel (web interface, API v1, MCP); the generators included in the interface depending on your plan are billed in tokens through the API: the table below is authoritative. This page lists everything, including what is free. The figures shown come directly from the application's billing configuration.
Three concepts, not one more
Credits
Credits are only used to create guides. Your plan grants a monthly allowance.
Tokens
Tokens pay for certain on-demand AI features. They are bought a la carte and stay valid for one year.
API quota
The API quota is the monthly cap on guides created through the API or MCP. It is independent from credits.
Credits: creating guides
Creating a guide costs the same number of credits everywhere: web interface, API v1 or MCP. Creating a guide never consumes tokens.
Cost per guide type
- SERP guide (Google): 1 credit(s)
- Google AI Overview guide: 1 credit(s)
- LLM guide (ChatGPT, Claude, Gemini, Perplexity, ...): 4 credit(s)
Monthly credit allowance per plan
| Plan | Credits per month |
|---|---|
| Scout Offer 🔦 | 1 credit(s) per month, non-cumulative: the allowance is topped up every month |
| Adventurer Offer 😎 | 4 credit(s) per month, non-cumulative: the allowance is topped up every month |
| Pioneer Offer 🤠 | 20 credits per month, cumulative up to 120 credits while the plan is active |
| Conqueror Offer 🚀 | Unlimited credits for manual use through the web interface (fair use). Through the API, the API quota applies. |
The credit is consumed when the guide is created: deleting a guide does not refund it.
Tokens: the master table, action by action
One row per action, one cost per channel. The MCP column always equals the API column: the MCP server is a plain relay to API v1, with no pricing of its own. Free rows are shown on purpose: you should see the zeros.
| Action | Web interface | API v1 | MCP |
|---|---|---|---|
| Create a guide (semantic analysis) | Credits only, no tokens | Credits only, no tokens | Credits only, no tokens |
| Compute or recompute a content score | Free | Free | Free |
| All reads: guide status, recommendations, balances, listings | Free | Free | Free |
| All deletions | Free | Free | Free |
| Search intent analysis | Included depending on your plan, no tokens | 15 tokens | 15 tokens |
| Internal linking suggestions | Included depending on your plan, no tokens | 15 tokens | 15 tokens |
| Outline generation | Included depending on your plan, no tokens | 12 tokens | 12 tokens |
| Meta tags generation | Included depending on your plan, no tokens | 8 tokens | 8 tokens |
| EEAT analysis of the content | 5 tokens | 5 tokens | 5 tokens |
| EEAT analysis of competitors | 5 tokens per analyzed competitor (maximum 10 competitors) | 5 tokens per analyzed competitor (maximum 10 competitors) | 5 tokens per analyzed competitor (maximum 10 competitors) |
| Content auto-optimization | 300 tokens per thousand words, prorated to the word count; applying an out-of-range iteration costs 50% of the initial cost | Not documented in the API (same token cost as in the interface) | Not documented in the API (same token cost as in the interface) |
| AIssistant (AI-assisted writing) | Variable depending on the chosen AI model | Not documented in the API (same token cost as in the interface) | Not documented in the API (same token cost as in the interface) |
Deliberate exception: the EEAT analysis consumes tokens in the web interface as well as through the API or MCP.
Concrete example: producing a full article with Claude through MCP
The typical protocol of an agent writing an optimized article, step by step:
- create_guides: start the semantic analysis of the query (1 credit(s))
- get_guide in a loop: wait for the analysis to finish (free, as many times as needed)
- Fetch the semantic recommendations (free)
- Write the content, in Claude or elsewhere (outside SERPmantics, nothing to pay)
- create_score: evaluate the content, iterate as much as needed (free)
- update_guide: save the final content (free)
Mandatory total: 1 credit(s) and 0 tokens.
Enriched variant: adding the optional generators through the API (search intent, outline, meta tags, internal linking) costs 50 tokens in total.
Buying tokens
Base price: 100 tokens for one euro excluding VAT.
- From 0 to 9,999 tokens: €1.20 excl. VAT per 100 tokens
- From 10,000 to 99,999 tokens: €1.00 excl. VAT per 100 tokens (base price)
- From 100,000 tokens upwards: €0.90 excl. VAT per 100 tokens (10 % discount)
Free purchase between 1,000 and 5,000,000 tokens.
Purchased tokens are valid for one year from the purchase date.
API quota and volume add-on
Plans with API access include 200 guides created per month through the API or MCP (monthly window anchored on your billing date).
Quota reached: the API answers with an explicit error (API_QUOTA_EXCEEDED). Nothing extra is ever charged behind your back.
Need more volume?
A volume add-on lets you target from 1,000 to 10,000 guides per month (in steps of 500). Only the guides beyond the 200 included ones are billed:
- Up to 4,500 guides per month: €0.18 excl. VAT per guide beyond the included quota
- Up to 9,500 guides per month: €0.17 excl. VAT per guide beyond the included quota
- Up to 10,000 guides per month: €0.16 excl. VAT per guide beyond the included quota
Frequently asked questions
Does checking a guide's status consume tokens?
+
No. All reads are free: guide status, recommendations, balances, listings. Your agent can poll the API in a loop without spending anything.
Does re-scoring a content several times cost anything?
+
No. Score computation is free, in the interface as well as through the API or MCP, as many times as needed.
Do tokens expire?
+
Yes, after one year: purchased tokens are valid for one year from the purchase date.
Is MCP more expensive than the API?
+
No. The MCP server is a direct relay to API v1: prices are strictly identical, action by action.
Does my plan grant me tokens every month?
+
No. No plan grants a monthly token allowance: tokens are only bought a la carte and stay valid for one year. Plans grant a monthly allowance of credits, used to create guides.
Can I limit my AI agent's consumption?
+
The operations that cost tokens are few and listed in the table above. Reads, scoring and deletions are free: an agent's iteration loops consume nothing. Only guide creation (credits) and the actions marked in tokens spend your balance.
Does creating a guide cost tokens?
+
No, never. Creating a guide only costs credits, at the same price in the web interface, through the API or through MCP.
Does deleting a guide refund the credit?
+
No. The credit is consumed when the analysis starts: deleting the guide does not give it back.
All prices are in euros excluding taxes (VAT applies on top).
© 2026 SERPmantics
Need help or have an idea? Send us an email at [email protected] !