Bezi

Name: Bezi
Brand: Bezi

AI Native

AI for building real games, not generating them.

Est 2021|Raised $13M|26 Ref Customers

LinkedIn 12K|Web traffic 43K|FTEs 27|HQ San Francisco, California, US

MarketsAI Coding Assistants

Published Oct 15, 2026

01Common complaints

- If you need broad, independently verified enterprise references - This product has limited visibility on mainstream review platforms based on this research session; independent enterprise references were not found. - If your primary need is IDE-only code completion - File-centric coding assistants such as Cursor or GitHub Copilot may be more cost-effective if in-editor Unity operations are not required. - If you are concerned about platform/vendor risk - If an organization prefers a first-party, engine-integrated solution, rapid improvements in Unity’s official assistants could reduce third-party value over time. - If your procurement requires fully predictable, flat pricing - Credit-based pricing introduces variability that some procurement teams may find difficult to budget for. - If your security posture requires openly published compliance artifacts - The vendor states SOC 2 is available by request, but ISO certifications and SOC 2 Type were not documented publicly in the sources reviewed. - Current limitations acknowledged by the vendor - Public documentation emphasizes the need for human review and control; explicit limitations (unsupported asset types, scale thresholds, Unity version compatibility) are not comprehensively enumerated in the public materials reviewed. - Features on the roadmap (not yet available) - The vendor blog highlights iterative additions (Actions, Workspaces, MCPs, Vision Tool) through early 2026, but no formal, dated public roadmap with commitments was found in the sources reviewed. - Scale limitations (at what volume/size does it break down?) - Scale characteristics are not publicly documented. - Security or compliance gaps - ISO certifications (e.g., ISO 27001/27017/27018) not publicly documented. - SOC 2 type (I vs II) not publicly documented. - Data residency options not detailed publicly. - Technical debt or architectural limitations mentioned in reviews - No substantial independent product review corpus was available; a single employee sentiment on Glassdoor referenced internal infrastructure improvement needs, which is not a product review. - API limitations or integration constraints - Specific API limitations and integration constraints were not clearly documented in the accessible sources.

Deal disqualifiers: what makes a POOR fit

Non-Unity engine environments (Unreal/Godot-first) where the product provides limited benefit.
Teams with strict data-residency or "no data leaves device" policies unless enterprise isolation options meet those requirements.
Buyers who require extensive third-party reviews or analyst validation prior to purchase and find the vendor's limited public review footprint insufficient.

Why some customers left for competitors (churn reasons inferred from community posts)

Community comments suggest:
- AI tools can be inconsistent during long sessions or when handling complex Unity-specific edge cases, requiring close human supervision.
- Some users prefer alternatives (e.g., Coplay) for their trial experience or different iteration workflows.

02Why they buy

Why customers chose this product over alternatives (based on vendor and community signals)

The vendor and community discussions highlight the following reasons users may prefer Bezi:
- The product is presented as providing Unity-specific context (assets, scenes, components) rather than only code-file context.
- In-editor operations that can modify scene objects and assets, with reviewable diffs and checkpoint/restore semantics.

03Field reports

Summary of review availability (important context)

- Major B2B review platforms (G2, TrustRadius, Gartner Peer Insights) - No accessible, product-specific Bezi review pages were found on TrustRadius or Gartner Peer Insights in this research session. - The Gartner Peer Insights site was accessible, but no Bezi-specific listings or reviews were found in accessible results. - A listing on OMR Reviews shows a Bezi entry with no reviews. - Resulting limitation - Most customer feedback in this report comes from vendor-published testimonials and community discussions on Reddit and Unity forums. These sources are not equivalent to independently validated enterprise references.

Vendor-published testimonials (examples, presented as vendor-supplied marketing content)

- A vendor-published testimonial attributed to Klas Holmlund (Creative Director) states that he finds Bezi valuable and subscribes to the service. - A vendor-published testimonial attributed to Nick Swift (SCH Games) states he uses Bezi for project-related questions rather than a general-purpose chatbot. - A vendor-published testimonial attributed to Nick De Paul (Wolf Soft Studios) reports that Bezi helped identify a bug quickly in his account; the claim is vendor-published and was not independently verified. - Note: These are vendor-supplied statements and are not independent, auditable measurements.

Community/social feedback — positive and negative signals

- Negative / churn signal - A Reddit thread titled “I just mass-deleted Bezi and Unity AI from my workflow. Here's why.” indicates some users have removed Bezi and Unity AI from their workflows, citing reasons such as a mismatch with expectations, cost, or workflow friction. These are anecdotal comments and not representative statistics. - Competitive pressure signal - Community commentary sometimes frames Bezi as a third-party plug-in and raises questions about potential platform risk if Unity’s first-party assistant improves. - Curiosity/awareness signal - Additional Reddit posts and forum threads indicate awareness and comparison discussions among users evaluating multiple AI tools for game development.

Specific outcomes customers achieved (with metrics when available)

Vendor testimonials include outcome claims such as time savings, for example reporting substantially faster module development or several weeks saved on debugging. These claims are anecdotal and vendor-published without an independently documented methodology.

Deal-breakers and reasons customers didn't purchase

Not well documented on review platforms due to the limited independent review corpus.
Potential concerns inferred from pricing and terms:
- Credit-based usage and variable costs can be a deterrent for buyers seeking predictable billing.
- Terms of service include binding individual arbitration for individual customers, which may be objectionable to some buyers.

Public case studies

The vendor site includes testimonial snippets but does not provide detailed, independently verifiable case studies (baseline → implementation → measured results) in the sources reviewed.

Vendor-published outcome claims (examples) and methodology gaps

Examples
- Testimonials include claims such as reduced development time and faster debugging.
Critical evaluation
- These claims are vendor-published and lack disclosed methodology, including measurement window, baselines, whether reported time includes review/QA, and sample sizes.

Mixed results / failed projects

Community content indicates some users have removed Bezi from workflows, but no structured, detailed post-mortem case studies were found in the reviewed sources.

04Best fit

Company sizes

Indie / small studios: very small teams where accelerated iteration and tooling can reduce staffing needs.
Mid-sized studios: small-to-mid sized teams where shared conventions, onboarding, and pipeline automation provide operational benefits.
Enterprise/AAA: potential fit when enterprise isolation, custom keys, SSO, and SOC 2 documentation satisfy internal security policies; typically subject to formal enterprise security review.

Industries and verticals where adoption is strongest

Game development using Unity.
XR / interactive 3D experiences developed in Unity (inferred from vendor testimonials and role references).

Roles and personas: decision makers AND end users

Decision makers: technical director, lead engineer, CTO, head of production.
End users: gameplay engineers, tools engineers, technical artists, designers working directly in Unity.

Pain points and challenges the product addresses

Repetitive Unity editor tasks (prefab setup, component configuration, scene wiring).
Debugging issues that cross code and scene-configuration boundaries.
Onboarding and consistency via stored conventions and documentation in Workspaces/Pages.

Qualification criteria: what makes a GOOD fit

Unity-centric workflows with significant in-editor work (as opposed to code-only workflows).
Teams prepared to perform human review of AI-generated changes and to maintain project conventions.
Use cases where automating Unity asset/configuration tasks can yield measurable time savings.

Real customer examples with company names when available

Vendor materials reference individuals with company affiliations; those references were not independently verified as paid customer relationships versus individual use or commentary.

05The pitch

- What the product does (core value proposition in objective terms) - Bezi is an AI assistant designed for Unity development. It indexes a Unity project (scripts, assets, scenes, packages) and uses that context to assist with coding, debugging, prototyping, and repetitive Unity-editor workflows. The vendor describes the product as intended for game creation rather than automated game generation. - A stated differentiation is that the tool operates inside the Unity Editor (rather than only in an external IDE): it can create/edit GameObjects, prefabs, components, materials, and automate content/pipeline tasks via features described as “Agent Mode” and “Actions.” These are vendor-described capabilities and should be validated in evaluation trials. - Target market and market positioning - Primary users: Unity developers, technical artists, designers, and small-to-mid game studios (including solo/indie developers). - Positioning: The vendor presents the product as a tool with project indexing and in-editor capabilities—terms used by the vendor include "project-aware" and "in-engine"; these vendor terms describe the product's intended scope (scene hierarchy + component awareness in addition to code files). - Company background (founding year, HQ, key leadership) - Leadership publicly referenced in vendor posts includes: - Julian Park (CEO) (author attribution on product posts). - Cecilia Uhr (Chief Product Officer) (author attribution on product posts). - Founding year / HQ: Not clearly documented on the public Bezi site pages and documentation reviewed for this report. - Funding and growth: investment rounds, acquisitions, estimated customer count/revenue if public - Funding - Bezi reported a Series A round in late 2023 in secondary press references; press items included additional details that were not independently verified for this report. - Growth / customers - Marketing pages state the product is "trusted by studios of all sizes" and invite users to join a large developer community, but no independently verifiable customer count or revenue was published in the sources reviewed for this report. - Acquisitions - No acquisitions were identified in the sources reviewed. - Pricing model with specific tiers and costs where available - Individual Plans (monthly) - Trial: Free with limited credits and constrained feature access (single workspace allowance and limited pages). - Pro: Published monthly plan with a moderate monthly credit allocation, limited workspace allowance, unlimited pages, access to all models; on-demand usage available at the vendor's published API pricing. - Advanced: Published monthly plan with a larger monthly credit allocation, expanded workspace allowance, unlimited pages, access to all models; on-demand at the vendor's published API pricing. - Ultimate: Published monthly top-tier plan with the largest monthly credit allocation and expanded workspace limits; on-demand at the vendor's published API pricing. - Studio Plans - Team: Published team pricing; includes features equivalent to the Ultimate tier plus shared monthly credits, Unity VCS & Perforce intelligence, unlimited shared workspaces/pages; a small number of seats are included with additional per-seat fees and additional credits for added seats. - Enterprise: Custom pricing; annual contracts only. - Credits - The vendor introduced model selection (Frontier/Core/Basic) and a revised credit system, including on-demand credits available at the vendor's published per-credit pricing. - Total cost of ownership considerations (implementation, training, hidden costs) - Cost drivers - Credits and overage/on-demand usage: heavy usage of higher-capacity models consumes more credits and drives variable costs. - Seats for studios: Team plan has per-seat fees and associated credit allocations. - Ongoing context/indexing and workflow-change management: teams may need to create and maintain Pages/Workspaces and conventions so the assistant remains aligned with project standards. - Implementation effort - Requires connecting Bezi to Unity projects and, for some workflows, integrating additional folders/tools (Workspaces, MCP integrations). - Potential hidden/indirect costs - Review and QA overhead: despite diffing and checkpointing features, teams must validate in-editor changes (scenes/prefabs/assets) to avoid regressions. - Security review effort for enterprise procurement: SOC 2 is stated by the vendor, but the report is provided by request via email rather than being self-serve publicly. - Key competitors in the space (list 4-6 direct alternatives) - Unity AI (Unity’s first-party AI assistant). - Coplay (Unity-focused in-editor assistant). - Cursor (AI coding IDE; not Unity-specific, typically code-file-centric). - GitHub Copilot (general coding assistant; not Unity-editor-aware). - Claude Code (general coding assistant; not Unity-editor-aware).

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

Download full report →