nunu.ai

Name: nunu.ai
Brand: nunu.ai

AI Native

Building the first multimodal gameplay agents.

Est 2022|Valuation $31M|Raised $8.5M|7 Ref Customers

LinkedIn 5K|Web traffic 5K|FTEs 24|HQ Zurich, Zurich, CH

MarketsTest Automation

Published Oct 15, 2026

01Common complaints

- When strong third-party validation is required prior to procurement - nunu.ai has a limited accessible presence on major review platforms, which increases perceived vendor risk for buyers that require extensive independent references. - When console support is required immediately - The vendor indicates console support is planned but not yet generally available; timelines and scope should be verified with the vendor. - When the QA strategy depends primarily on code-level assertions and engine-level instrumentation - Black-box, agent-driven testing does not replace unit/integration tests or engine-instrumented harnesses for code-level verification. - When there is no capacity for test triage - Agent-driven failures require human review to interpret recordings and determine whether failures reflect defects or agent limitations. - When strict cost predictability is essential - Credits consumption and retry behavior can create variable spend; fixed-license models or in-house automation may offer more predictable costs. - Limited third-party review footprint - A small public review presence makes it harder to benchmark reliability, flakiness rates, and support responsiveness. - Console support availability - Console support is indicated as planned; current availability and production readiness should be confirmed. - Public security documentation depth - SOC 2 Type II is stated, but public artifacts such as subprocessor lists, audit logging details, SCIM support, and detailed data residency controls were not clearly documented in the reviewed pages. - Operational limits implied by pricing - Starter plan device availability is subject to availability, suggesting capacity constraints for free-tier users. - Self-serve concurrency begins at 1 device; scaling parallelism will affect consumption and cost.

02Why they buy

- Vendor-claimed differentiators (objective description) - AI agents intended to operate across game and app UIs on real devices, interacting via screen observation and input simulation. - Natural-language test authoring to reduce the need for scripted test code. - Access to a real-device execution pool (iOS/Android) and options for Windows devices via custom arrangements. - Evidence caveat - Vendor marketing statements (for example, quantified savings) are presented without independent validation in the public materials reviewed; buyers should validate claims through pilots and reference checks. - Where competitors may have the advantage - Larger incumbents may provide more mature compliance documentation, audit trails, and a larger set of independent customer references.

03Field reports

Customer Testimonials

- Availability of independent reviews - Major B2B review platforms (G2, TrustRadius, Gartner Peer Insights) did not surface an accessible nunu.ai product page with user-generated ratings/reviews in the searches performed; this may reflect a limited review footprint or discoverability issues. - Vendor-published testimonials/case-study claims - A vendor-published case study from Roboto Games reports a notable reduction in QA costs and a very short integration time; it also cites automated notifications and recordings/reports for failed tests. These outcomes were published by the vendor and the methodology or baselines are not disclosed. - Community discussion signals - Threads on game development and QA forums reference automated QA for games and mention nunu.ai in the context of a still-developing space. Community commentary indicates skepticism about the maturity of AI-driven QA for all edge cases and notes that many teams continue to rely on human QA and engine-level testing for certain scenarios. - What is missing - There is limited public, third-party, long-form buyer feedback on detailed reliability metrics (e.g., flakiness, false positives/negatives, or support responsiveness). This increases buyer risk and makes a structured pilot or proof-of-concept (POC) advisable.

Roboto Games / Stormforge (vendor case study)

Reported outcomes: a notable reduction in QA costs; a short integration time; faster automated testing with notifications and recordings.
Critical evaluation: Metrics are vendor-published without disclosed methodology, sample size, measurement period, or counterfactual. Buyers should request a breakdown showing baseline QA cost model, time window measured, defect leakage impact, and categorization of failures.

Vendor technical content describing black-box testing for games and apps

The vendor documentation describes recordings, logs, issue annotations, and general UI navigation across mobile apps and PC applications. This material illustrates intended functionality but does not constitute independent validation of outcomes.

04Best fit

Ideal Customer Profile

- Company sizes - Likely best fit: - Small-to-mid-sized game studios and mobile app teams that have recurring manual regression testing burdens. - Heuristic: small-to-mid-sized organizations or teams with a small QA/QE staff where regression testing is a recurring bottleneck. - Industries and verticals - Primary: games (mobile and PC studios), live-ops teams, QA outsourcing managers. - Secondary: consumer-facing applications with complex UI flows and frequent UI changes. - Roles and personas - Decision makers: - Head of QA / QA Director - VP Engineering / CTO (automation strategy) - Production leadership responsible for release cadence and cost considerations - End users: - QA engineers - Release engineers / CI owners - SDETs for integration and guardrails - Pain points the product addresses - High manual regression workload. - Difficulty automating game UI flows with traditional frameworks. - Need to run scheduled or wide-coverage tests on real devices. - Qualification criteria (good fit) - Presence of repeatable flows (tutorials, purchases, onboarding, menus) that are important for revenue or retention. - Operational capacity to triage test results and distinguish agent limitations from real defects. - Ability to run a structured pilot with defined acceptance criteria. - Deal disqualifiers (poor fit) - Requirements for fully deterministic, identical runs for compliance-grade regression without exceptions. - Strict security or regulatory requirements that necessitate extensive public security documentation prior to any pilot. - Products with heavy anti-bot mechanisms that prevent agent interactions.

05The pitch

What the product does (core value proposition in objective terms)

- nunu.ai provides AI agents that execute end-to-end black-box QA tasks on real devices by observing the screen and interacting via taps/clicks/typing, based on natural-language instructions (e.g., “complete the tutorial,” “finish signup,” “make a purchase and confirm receipt”). - The product is positioned to reduce manual QA effort, expand test coverage, and enable scheduled or continuous test execution; it provides recordings and reports intended to assist with debugging.

Target market and market positioning

Primary: game studios (mobile and PC) seeking automated gameplay/UI regression testing and broader QA coverage.
Secondary: application teams with end-to-end UI workflows that are brittle under traditional scripted automation.
The vendor promotes a no-code, natural-language test authoring model and operation on real devices.

Get a comprehensive analysis of nunu.ai including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →