Everything you need to run production AI
One OpenAI-compatible gateway for routing, reliability, cost control, access, safety and observability across the models you use.
Routing and reliability
Send each request to the right model and provider, and keep it answering when one fails.
Provider routing
Send the same model to the best available provider, and move traffic when one fails.
Fallback models
Name a backup model, and Spro tries it when the first model cannot answer.
Virtual model aliases
One stable model name in your code, pointed at any model or provider behind the scenes.
Automatic model selection
Let the gateway choose the model for each request, based on the task, cost and quality rules you set.
Cost control
Lower the bill with cheaper tiers, caches and batch work, and see exactly where every token goes.
Flex pricing
Run the same model at a lower price on spare provider capacity, for work that can wait.
Response caching
Return a saved answer for repeated or similar requests, without calling the model.
Prompt caching
Reuse the provider's processed prompt prefix to cut time to first token and input cost.
Prompt compression
Shorten long prompts at the gateway so you pay for fewer input tokens.
Batch processing
Send many chat or embedding requests in one file, and collect the results when the batch finishes.
Spend tracking
See tokens, cost and spend per request, model, provider, key and project.
Usage analytics
See AI traffic as trends by model, provider, key and team, with tokens, errors and cost over time.
Cost attribution
Tie every AI charge to the key, model and provider that caused it, so every dollar has an owner.
Key spend limits
Set a budget on each API key, and get warned before it runs out.
Token allocation
Assign token budgets to teams, projects, apps and keys before usage grows.
Centralized billing
See your balance, key usage, purchases and transactions in one billing view.
Volume-based pricing
Commit to a token volume up front and get a volume discount, with a platform fee that does not rise as you grow within a plan.
Access and keys
One key for many providers, your own provider keys, and separate workspaces for each team.
Universal API key
One Spro key and one endpoint for models from OpenAI, Anthropic and other providers.
BYOK
Use your own provider keys while Spro handles routing, guardrails, caching and tracing.
Provisioning keys
Create, rotate and revoke API keys from code, instead of by hand in the dashboard.
MCP gateway
Register tool servers once and make their tools available to every model you route.
Team workspaces
Separate workspaces for each team or product, so keys, guardrails and spend stay with the team that owns them.
SSO and SAML
Require sign-in through your identity provider, and add people to the right organization automatically.
Safety and governance
Check every prompt and response, protect personal data, limit traffic and control who can do what.
Guardrails
Check every prompt and response, then log, warn, block or redact what breaks your rules.
Data redaction
Find personal and confidential values in prompts and replace them with placeholders before the model sees them.
Rate limits
Cap requests and tokens per minute, hour or day, so one app cannot flood a provider or your budget.
Zero data retention
Spro does not store the content of your requests or responses in a workspace where ZDR is on.
RBAC
Organization roles that decide who can see, change and manage what in your workspace.
Audit logs
A record of who changed members, teams, keys and settings, and when.
Model access controls
Choose which models each API key can call, so a key only reaches the models it is meant for.
Observability
Know what every request did, how long it took, and get told when spend, errors or models change.
Request logs
Every request with its model, provider, tokens, latency and trace ID, in one searchable log.
Tracing
Follow one request, or a whole agent run, through Spro with trace IDs and resolved-model headers.
Latency tracking
See how long each request takes by model, provider and key, so slow routes show up before users complain.
Alerts
Get notified in the app and by email when errors spike, balance runs low or a model changes.
Dynamic tags
Attach labels to each request so you can slice usage and cost by feature, customer or team.
Insights
Get evidence-based cost and performance recommendations from your own traffic.
Spend spike alerts
Get notified when spend breaks from its normal pattern, with sensitivity you control.
Error rate alerts
Get an email when the share of failed requests passes a level you set, for the owner of that traffic.
Balance alerts
Get warned when your prepaid balance drops below a threshold, or is set to run out soon.
Model usage reports
Get a scheduled email summary of spend, requests and tokens by provider and model.
Quality
Measure answers on your own data and improve the prompts behind them.
Evaluations
Compare models on your own data, scored by an LLM judge or by criteria you define.
Prompt optimization
Improve prompts by testing variations against your criteria and keeping the ones that score better.
Prompt management
Version your prompts, compare versions side by side and roll back when a change hurts.
Route every AI request through SPRO.
One OpenAI-compatible API for 350+ models, with intelligent routing, 29 built-in guardrails, observability, and automatic failover.
