Respan

Respan is an AI router with built in observability and automated evaluation. It gives developers one API endpoint that reaches more than 1,000 models across major providers, then traces, monitors and scores every agent run that passes through it. Teams point the SDK they already use at the Respan gateway and switch models by changing a single value. The gateway handles automatic fallbacks when a model errors or is rate limited, response caching for repeated requests, load balancing, a key vault for bringing your own provider keys, and spending and rate limits at the key, customer or organization level. The observability side covers logs, agent tracing, cost and usage tracking, alerts, custom dashboards and OpenTelemetry support. Evaluations combine LLM judges, deterministic code checks and human review queues, with test cases sampled from production traffic or uploaded as CSV. Prompt management and red teaming are also included. The Free plan costs $0 and includes 100k logs, 1k scores, 5 datasets, 2 evaluators and 5 prompts. The Team plan is $199 per month billed yearly and adds unlimited datasets, evaluators and prompts, a private Slack channel and a SOC 2 report, with five members included and extra seats at $15 per member. Additional logs are $8 per 100k and additional scores are $1 per 1k. Enterprise is custom, with volume discounts, custom SLAs, self hosting and a HIPAA BAA. A HIPAA add on is listed at $249 per month.