Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.responsibleailabs.ai/llms.txt

Use this file to discover all available pages before exploring further.

Every API response includes a credits_consumed field showing the exact cost. Cached results always cost 0 credits.
Credits reset monthly and do not carry forward. Upgrade your plan or contact support for a top-up if you need more.

​
Credit costs by endpoint

​
Evaluation - POST /railscore/v1/eval

ModeDimensionsCredits
BasicAll 81.0
BasicSingle dimension0.2
BasicCustom (n dimensions)min(0.3 × n, 1.0)
DeepAll 83.0
DeepSingle dimension1.0
DeepCustom (n dimensions)min(2.0 × n, 3.0)

​
Safe Regeneration - POST /railscore/v1/safe-regenerate

StepCredits
Each evaluation (basic mode)1.0
Each evaluation (deep mode)3.0
Each regenerationsame as eval mode
Total = eval_credit × (evaluations + regenerations). Up to 5 iterations per request.

​
Compliance - POST /railscore/v1/compliance/check

ScopeCredits
Single framework5.0
Multi-framework (2–3)8.0
Multi-framework (4–5)10.0

​
Agent endpoints

EndpointCredits
POST /railscore/v1/agent/tool-call1.5 (standard) / 3.0 (deep)
POST /railscore/v1/agent/tool-result0.5–1.0
POST /railscore/v1/agent/prompt-injection0.5

​
Free endpoints

GET /health and POST /verify always cost 0 credits.
Identical requests within the cache window return cached results at 0 credits. Basic mode caches for 5 min, deep mode for 3 min, compliance for 5 min.

​
Plan allocations

PlanCredits / month
Free100
Pro1,000
Business10,000
EnterpriseCustom

​
Requests per plan

OperationCostFree (100)Pro (1,000)Business (10,000)
Basic eval, all 8 dimensions1.01001,00010,000
Deep eval, all 8 dimensions3.0333333,333
Basic eval, single dimension0.25005,00050,000
Compliance, single framework5.0202002,000
Safe-regen, basic (3 iterations)3.0333333,333

​
Monitoring usage

Track credit consumption in real time on your dashboard. The Credit History section shows usage milestones at 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 85%, 90%, 95%, and 100% of your monthly allocation. You can also check per-request cost programmatically. Every API response includes credits_consumed: 
result = client.eval(content="...", mode="basic")
print(result.credits_consumed)  # e.g. 1.0

​
Important notes

  • Credits reset monthly and do not carry forward to the next period.
  • Insufficient credits returns HTTP 402 with your current balance and the required amount.
  • Top-ups are available on request. Contact support or upgrade your plan.

​
What’s next

Quickstart

Make your first scored request in under 5 minutes.

Authentication

Set up your API key and manage environments.

Evaluation API

Full parameter reference for the eval endpoint.

Dashboard

Monitor credit usage in real time.