Free
for trying the API and shipping a first integration
$0 always free
- $5.00 signup credit
- 1 workspace seat
- Public access without an account
- Fast + Smart classification
pricing
simple plans with upfront usage
Pro
for developers and personal agents in production
$20 / month
- $20.00 of usage each month
- 3 workspace seats
- 10× rate limits
- Usage by connection and agent
Enterprise
for teams that need capacity, infrastructure or a contract
Custom
- Volume-based capacity
- Dedicated deployment
- Private inference options
- Measured accuracy on your data
Usage prices
Scrape and classify a public URL with a funded workspace key. Request Markdown and HTML at no extra cost. Provider-billed scrapes remain charged if classification fails; errors disclose the charge. See the URL API.
| Usage | Price |
|---|---|
| URL scraping | $2.20 / 1,000 requests + classification |
| Input tokens | $0.042 / million |
| Smart escalation | +$2.00 / 1,000 |
| Jev long context (original context tokens) | $0.084 / million |
Smart starts with Fast and reviews uncertain answers. You pay the extra charge only for answers that are successfully reviewed. No escalation means no extra charge.
For example, 1 million input tokens with 50 Smart escalations cost $0.142.
Output tokens are free. Input usage includes the text, labels and instructions processed by the base classifier. Retries, fallback routing and Smart model tokens add no separate charges. Paid requests already admitted can finish and leave a negative balance. New requests require a positive available balance. No automatic top-ups.
Jev long context starts automatically above 32,000 characters with the default model or explicit Jev. It costs 2 × Jev's $0.042 rate: $0.084 per million original context tokens, counted with cl100k_base once across inputs. Dimensions and actual screening or final-call usage do not multiply this price. 250,000 context tokens cost $0.021.
Requires paid workspace balance or an active paid subscription; anonymous access and free signup credit do not qualify. Fast only. Synchronous requests allow up to 250,000 original context tokens, 20 documents and 32 decisions within 1 MB. Each document × dimension or multi-label category counts as a decision.
Upload one whole document of up to 10 million original tokens (100 MB) at the same rate. Splitting and screening happen automatically. A full job costs $0.84. The workspace reserves the actual uploaded document's token price, then settles once when final judgment succeeds. Jobs expire after 24 hours; failed or canceled jobs are refunded.
Final Jev reads selected evidence. Eligible chunks can be omitted when the final budget fills; usage.long_context reports selection. No usable evidence returns 422 long_context_no_evidence without charge. Explicit chunklaya remains a separate legacy opt-in. Read the long-context limits.
Included on every plan
Fast and Smart classification with your own labels through REST, MCP or the CLI.
Rate limits
| Plan | Fast | Smart |
|---|---|---|
| Free | 3,000/min · 20,000/day | 200/min · 2,000/day |
| Pro | 30,000/min · 200,000/day | 2,000/min · 20,000/day |
Limits count classifications and are shared across workspace keys and agents. Public access is limited per IP. Laya trial limits apply to every plan.
Free access allows up to $0.01 of provider cost per Fast request or $0.10 per Smart request, with $0.50 per IP per UTC day and four requests at once. Smart reviews beyond the request cap retain their Fast answers. A funded workspace has its own allowance and supports larger Smart requests. Signup credit alone uses the free limits.
See request limits and retry guidance. Usage is reserved before a request and unused funds are released afterward. Anonymous proxy networks require a funded key.
