Home / Directory / AI SaaS tooling / Observability & monitoring / Respan
Respan
Respan, formerly Keywords AI, is a San Francisco software company whose platform routes calls to more than 1,000 models through one gateway and traces, monitors, and evaluates AI agents in production.
Startups and product teams shipping AI agents that want one API for model access plus tracing, alerts, and evaluations, with a free plan to start and a team plan at $199 a month.
Teams that want an observability product with a long track record or a large vendor behind it, and buyers who need on-premises deployment spelled out before a call.
Verdict
Respan is worth a look when a team wants model routing and observability from one vendor. It posts a free plan and a $199 monthly team plan with per-log overage prices, lists ISO 27001, SOC 2, GDPR, and HIPAA with a business associate agreement, and ships an open-source Apache-2.0 gateway. It is a young company with a $5 million seed round, so buyers should weigh that against the larger vendors in this field.
Score Breakdown
These scores cover how well Respan traces and debugs agents, scores outputs, routes and falls back across models, and tracks cost and alerts, and then how it does as a company on innovation, communication, pricing, and contract terms.
Who Uses Respan, Where It Fits on Price, and Its Competitors
Respan puts one API in front of more than 1,000 models, with fallbacks, caching, and budgets, and records every call as a trace so teams can see what each step of an agent ran and returned. On top of that it adds monitoring, alerts, and evaluations that score outputs. Respan says it processes more than a billion logs a month, and its homepage shows testimonials from Retell AI, Lovable, and Mem0. The company was called Keywords AI until it renamed itself, and its gateway code is open source under Apache-2.0.
Respan has a free plan, a Team plan at $199 a month, and a custom Enterprise plan. Its closest competitors are Fiddler AI, Laminar, Langfuse, and Raindrop, which score slightly higher in our reviews, Arize AI, which scores just behind, and Evidently AI, which scores lower.
Pricing
We checked Respan's pricing page and recorded the free plan, the Team plan with its usage rates, and the Enterprise plan.
| Plan | Price | What stands out |
|---|---|---|
| Free | $0 | Entry plan for trying the gateway and tracing. |
| Team | $199 a month | Includes 100,000 logs. Extra logs cost $8 per 100,000 and scores $1 per 1,000, and seats cost $15 per member. |
| Enterprise | Custom | Adds negotiated terms. A HIPAA add-on is listed at $249 a month. |
| Security | Listed | ISO 27001, SOC 2, GDPR, and HIPAA with a business associate agreement, according to Respan. |
The Field at a Glance
How Respan compares with other observability and monitoring vendors we reviewed, by Overall Score and relative typical engagement cost.
Respan scores 7.6 in observability and monitoring, under Fiddler AI (8.1), Laminar (7.9), Langfuse (7.9), Raindrop (7.9), and Arize AI (7.7), and ahead of Empirik (7.2), Evidently AI (6.8), and Deepchecks (6.7).
Compared with …
- Respan vs Langfuse 7.6/7.9
How Well Respan Fits Each AI Observability Use Case
We rated Respan against the situations product teams bring to an observability platform, drawing on its homepage, pricing and about pages, Y Combinator profile, and March 2026 funding announcements.
| Use case | Fit | Notes |
|---|---|---|
| Debugging a failing agent run | Strong | Every span records the exact input and output, and Retell's CTO credits Respan with resolving production issues 10 times faster. |
| Routing and fallbacks across providers | Strong | One API reaches more than 1,000 models with fallbacks, caching, and budgets. |
| Scoring outputs and catching regressions | Strong | Evaluations and scoring are built in, with scores priced at $1 per 1,000. |
| Healthcare workloads | Mixed | HIPAA with a business associate agreement is listed, and a HIPAA add-on is priced at $249 a month. |
| Self-hosting the whole platform | Weak | The gateway code is open source, and the observability platform is a hosted service. |
| Buyers who want a long track record | Mixed | The company is a Y Combinator Winter 2024 startup that raised a $5 million seed round in March 2026. |
Review Excerpts
These excerpts come from testimonials Respan posts on its homepage, which the company chose, and from the buyer concern we found on its pricing page. Paid reviews and pay-for-play sites were excluded.
“We scaled from 5M to 500M+ monthly API calls quickly. Respan gave us the debugging layer to resolve production issues 10x faster.”
“Respan has been key in helping us scale to trillions of tokens reliably with real-time observability.”
“The Team plan is clear at $199, but logs, scores, and seats each bill separately, and I can't tell what Enterprise adds beyond the HIPAA add-on.”
Where customer reviews couldn't be found, our analysts attempted to give a voice to buyer sentiment, drawing on real information from review sites, forums, and product documentation.
Based on: Respan pricing
Company Facts
- Founders
- Andy Li (CEO), Raymond Huang (CTO), and Hendrix Liu (head of sales) (Respan about)
- Y Combinator
- Winter 2024 batch (Y Combinator)
- Headquarters
- San Francisco (Y Combinator)
- Funding
- A $5 million round from Gradient, Y Combinator, and other investors, announced March 18, 2026 (GlobeNewswire)
- Products
- AI gateway, tracing, monitoring, and evaluations (Respan)
- Open source
- Apache-2.0 gateway on GitHub (GitHub)
- Website
- respan.ai
Key Leadership
Andy Li
Co-founder and CEO
Respan's about page and Y Combinator profile list him as chief executive, and he previously worked as a product design engineer at Apple (Y Combinator).
Raymond Huang
Co-founder and CTO
His Y Combinator profile lists a mechanical engineering degree from the University of Illinois Urbana-Champaign (Y Combinator).
Hendrix Liu
Co-founder and head of sales
Respan's about page lists him as head of sales (Respan about).
Methodology
We scored Respan on eight criteria, from tracing and debugging to contract fairness, and the Overall Score is their simple average on a 0 to 10 scale. The scores draw on reviews from public review sites and forums, with paid reviews and pay-for-play sites excluded, plus a hands-on look at the product, its pricing, and its documentation and training. Scores can change as we add reviews and product checks.