How a list becomes structured data.
Paste a list. Toolsy pulls out item names, quantities, units, and specs. You see every line before an RFQ draft is prepared.
From a site note to a quote package.
Each step can be checked on its own. Without an API key, the parser runs alone.
Paste
Paste a list or upload a text file.
Read
Toolsy pulls out item names, quantities, units, and specs.
Check
You see every line before an RFQ draft is prepared.
Match
Each line is compared with a reference directory of 67 catalog items.
Draft
One RFQ email per supplier, ready to send from your own mailbox.
The rules the model works under.
When the model output does not fit, the workspace falls back to the parser. It does not guess to fill a gap.
Rules take over when needed
If the model output does not fit, the parser result is used.
Source lines stay attached
Every extracted line keeps the original text beside it, so you can compare the two.
You choose the supplier
Toolsy never picks a winner. It ranks listings and shows the trade-offs; you select the supplier.
No training on your text
Toolsy does not use workspace text to train models. Your model provider's terms still apply.
Read each line before you send.
Each extracted line keeps the source text it came from, so you can compare the two side by side.
300 pcs 4-inch PVC pipe AW class
- Name
- PVC pipe
- Spec
- 4-inch · AW class
- Quantity
- 300 pcs
- Category
- Plumbing & Pipe
For developers
Any OpenAI-compatible endpoint works. The provider and model are set per deployment.
| Task | Volume | Use |
|---|---|---|
| Material-list extraction | One model pass per list, when configured | Schema-validated JSON output |
| RFQ draft rewrite | Optional, one pass per draft | Plain-text subject and body; template fallback |
| Catalog matching | Reference catalog, refreshed by hand | Name and alias matching |
Next.js · Postgres · OpenAI-compatible inference · [email protected]
Try your own list.
The free plan has no card requirement.