Patronus AI
Patronus AI, Inc.
AI evaluation company turned research lab selling training data and simulated environments to frontier AI labs; its eval API is still documented
Last checked October 15, 2026Quick facts: Patronus AI is an AI compliance vendor founded in 2023 and headquartered in San Francisco, United States. Pricing is available on request. Profile last checked October 15, 2026; sources are cited at the bottom of this page.
About Patronus AI
Patronus AI was founded in 2023 by Anand Kannappan (CEO) and Rebecca Qian (CTO), who both previously worked on AI at Meta, and is based in San Francisco. It launched in September 2023 as an LLM evaluation company for regulated industries. It raised a $3 million seed round led by Lightspeed Venture Partners, then a $17 million Series A led by Notable Capital in May 2024. In June 2026 it raised a $50 million Series B led by Greenfield Partners, with Lightspeed, Notable Capital, Datadog and Samsung, bringing its total to $70 million. **What it sells now** Its homepage now says: "Patronus AI is a research lab developing training data and simulated environments for frontier AI." The products page offers Frontier Training Data: tasks with reference solutions and graders, and ready-built or custom simulated environments for software development, computer use and knowledge work, delivered to run in a lab's own training infrastructure. In December 2025 it launched Generative Simulators and RL Environments, which generate tasks and rules and adapt them as an agent learns. In June 2026 it announced Digital World Models, replicas of websites and company systems where agents can practice. Patronus says most leading frontier labs and hyperscalers are customers but names none, and it reports revenue growth of more than 15 times over the past year. **The evaluation platform** Patronus's documentation still describes the evaluation product the company started with. It includes an API and SDK with off-the-shelf evaluators (among them Lynx for hallucination detection and the small Glider evaluator model), custom evaluators, experiments, logs, tracing and Percival, launched in May 2025 to find failures in agentic systems and suggest fixes. Evaluation Runs and the Experimentation SDK are Enterprise features. An older landing page names OpenAI, HP and Pearson as customers and offers on-premise deployment. The homepage, products page and pricing page no longer mention the platform. **AI compliance coverage** Patronus doesn't map its products to any AI regulation or standard and publishes no security certifications. Its older landing page says it is vetted by third-party security firms each year. **Who it's for** Frontier AI labs and large AI teams that buy training data and agent environments. Teams looking for an evaluation or observability tool to support compliance work should confirm that the evaluation platform is still sold and supported before buying.
Framework coverage
Patronus AI features
Capabilities Patronus AI markets publicly. Inclusion means the feature is documented on the vendor's site — not that it's best-in-class. Last checked October 15, 2026.
LLM Evaluation
Systematic testing of LLM outputs for correctness, relevance, safety, and consistency using automated scorers, rubrics, or human review.
Hallucination Detection
Automated scoring of LLM outputs for factual accuracy and groundedness, typically via retrieval-backed verification or reference comparison.
Agent Tracing
End-to-end visibility into multi-step LLM agent runs: tool calls, intermediate reasoning, token usage, latency, and errors at each step.
LLM Observability
Logging, tracing, and metrics for LLM applications in production — latency, cost, quality, and error rates across traces and users.
Patronus AI pricing
No published prices. The pricing link on patronus.ai now goes to an archived page that asks buyers to contact the company about 'training needs, project scope, and pricing'. The earlier developer credits and per-call evaluator prices are no longer listed. The documentation marks Evaluation Runs and the Experimentation SDK as Enterprise features.
Third-party reviews
Publicly cited ratings from independent review platforms. We don't collect reviews ourselves and we don't weight these in our rankings. Low review counts (≤2) are noted so you can decide how much signal to take from them.
G2
12 reviews
We publish the cited numbers as-is and do not combine them into an aggregate rating, on this page or in search-result markup.
Vendor changelog
Durable events we've independently sourced — acquisitions, certification milestones, product launches. Each entry cites a primary or trusted secondary source.
- funding
Raised $50M Series B led by Greenfield Partners
With Lightspeed Venture Partners, Notable Capital, Datadog and Samsung; total raised reached $70M. Announced alongside Digital World Models.
SiliconANGLE - product launch
Launched Generative Simulators and RL Environments
Adaptive training environments for AI agents, sold to foundation model labs and enterprises.
VentureBeat - product launch
Launched Percival
A self-serve tool that finds failures in agentic systems and suggests fixes.
PR Newswire - funding
Raised $17M Series A led by Notable Capital
With Lightspeed Venture Partners, Factorial Capital and Datadog.
TechCrunch
Pros and cons of Patronus AI
Pros
- Long research record on LLM and agent evaluation (FinanceBench, Lynx, Glider, TRAIL), and the evaluation API and SDK are still documented.
- Percival finds failures in agentic systems and suggests fixes; Patronus says it detects more than 20 failure modes, such as incorrect tool use, context misunderstanding and planning errors.
- Well funded: $70M raised, including a $50M Series B in June 2026.
- G2 users rate it 4.3 out of 5 from 12 reviews (October 2026).
Cons
- The company now sells training data and simulated environments to frontier AI labs. The evaluation platform is no longer on its homepage, products page or pricing page, so confirm that it is still sold and supported.
- No published prices: the pricing page is archived and asks buyers to book a call.
- No documented mapping to any AI regulation or standard, and no published SOC 2 or ISO 27001 attestation.
- Few named customers: its current pages name none, and only an older landing page lists OpenAI, HP and Pearson.
Alternatives to Patronus AI
Frequently asked
What is Patronus AI?+
Patronus AI is a San Francisco company founded in 2023 by Anand Kannappan and Rebecca Qian, who previously worked on AI at Meta. It began as an LLM evaluation platform. It now describes itself as a research lab developing training data and simulated environments for frontier AI, although its evaluation API and SDK are still documented.
Does Patronus AI still sell an LLM evaluation platform?+
Its documentation still covers evaluators such as Lynx and Glider, experiments, tracing and Percival, but its homepage, products page and pricing page no longer mention them. Ask Patronus whether the evaluation platform is still offered to new customers before you plan around it.
How much does Patronus AI cost?+
Patronus doesn't publish prices. Its pricing page is archived and asks buyers to contact the company about their training needs and project scope. The earlier free developer credits and per-call evaluator prices are no longer listed.
How much funding has Patronus AI raised?+
$70 million in total: a $3 million seed round led by Lightspeed Venture Partners (2023), a $17 million Series A led by Notable Capital (May 2024) and a $50 million Series B led by Greenfield Partners (June 2026).
Has Patronus AI been acquired?+
No. We found no acquisition of Patronus AI as of October 2026. It raised a $50 million Series B in June 2026 and changed its focus from evaluation tools to training data and simulated environments for frontier AI labs.
Sources
- Patronus AI: Homepage (research lab positioning, October 2026)
- Patronus AI: Products (Frontier Training Data: tasks, environments, research)
- Patronus AI: Pricing (archived; contact for pricing)
- Patronus AI: Company (founders)
- Patronus AI docs: Evaluation platform overview (evaluators, Lynx, Glider, Percival, tracing)
- Patronus AI docs: Core concepts (Enterprise features)
- Patronus AI: Older evaluation landing page (customers, on-premise option)
- Patronus AI: Press (Glider, Judge-Image, Generative Simulators, Series B)
- Patronus AI press release: $50M Series B and Digital World Models (June 26, 2026)
- The Next Web: Patronus AI raises $50M Series B (June 2026; San Francisco, $70M total)
- SiliconANGLE: Patronus AI raises $50M for world models (June 25, 2026)
- Notable Capital: Why we're doubling down on Patronus AI (June 2026)
- VentureBeat: Patronus AI launches Generative Simulators and RL Environments (December 17, 2025)
- PR Newswire: Patronus AI unveils Percival (May 14, 2025)
- TechCrunch: Patronus AI raises $17M Series A led by Notable Capital (May 2024)
- TechCrunch: Patronus AI launches with $3M seed (September 2023)
- G2: Patronus AI reviews
Keep reading
See an error or outdated detail?
Profiles carry a last-checked date. If something is out of date or wrong, send a correction and we will review it.
Work at Patronus AI?
Claim this listing to propose edits to the tagline, description, pricing notes, and headquarters details. Every change is still reviewed by our editorial team.