Quick Verdict
Bijian Studio, written 必剪Studio in Chinese, is Bilibili’s active web product for digital avatars and synthetic voice video. Its official entry remained reachable on July 21, 2026, but the workspace and current entitlements require a Bilibili account. The old description of a free closed beta and waitlist is obsolete. A more accurate status is an account-gated product with free and VIP quotas that can change by account, campaign, and product stage. No fixed allowance or permanent-free promise should be preserved.
It can fit a Bilibili creator who operates a real personal brand and has explicitly authorized use of their own face and voice for scripted videos. It should not be used to make a public figure, employee, customer, friend, or family member appear to say something without consent. Face images, voiceprints, identity records, source footage, and derived models are highly sensitive. The logged-out page does not resolve product-specific retention, training, model ownership, deletion verification, and output-rights questions. Until those terms are reviewed inside the account or confirmed in writing, uploading irreplaceable biometric material is not recommended.
Best For
- Established Bilibili creators producing repeatable, low-risk scripts under their own identity.
- A person creating a digital avatar from their own face and voice with informed consent.
- Teams that can document the account owner, represented person, script approver, and publisher.
- Creators willing to review every generated video and disclose synthetic presentation.
- Not suitable for anonymous endorsements, unauthorized cloning, deceptive evidence, regulated advice, or teams without a model-deletion process.
Teams that need stock avatars rather than a clone of a real operator can compare HeyGen and Synthesia. Developers building real-time face-to-face agents can review Tavus.
Key Features
- Digital avatar creation: source footage meeting current requirements can be used to create a talking virtual likeness.
- Voice capability: recording or uploaded speech can support a voice associated with the avatar; exact capture rules are shown in the workspace.
- Script-to-video generation: text can be rendered as a talking-avatar video, reducing repetitive filming.
- Bilibili account integration: login, identity, entitlements, and publishing connect to the Bilibili creator ecosystem.
- Free and VIP quotas: account classes can receive different generation counts, duration, queueing, templates, or related benefits.
- Browser workspace: creation is online, so upload, rendering, and export depend on the live service.
Use Cases
For an owned creator channel, the represented person can authorize their face and voice, approve each script, and inspect pronunciation, expression, lip sync, and visual artifacts before publication. For course updates, restrict the avatar to low-risk supplementary material and retain a real instructor contact and correction process. Short versions of an intro or recurring explanation may save filming time, but should not be presented as live footage or eyewitness evidence.
An organization needs separate, informed, revocable written consent from the represented employee. The agreement should cover channels, term, editing, advertising, sublicensing, withdrawal, departure from employment, takedown, and deletion of source media and derived models. Operations staff should never build an avatar from old employee videos without that permission. Do not use the product for fraud, fake endorsements, identity bypass, unauthorized voice cloning, or misleading medical, financial, legal, government, or news statements.
Pricing
Bijian Studio should be classified as freemium, not permanently free. Free and VIP account states may carry different generation counts, duration, queue priority, templates, or other entitlements. Those values can change with promotions and product updates. The logged-out public page is not a reliable fixed price table, and the older free-beta description no longer defines the current offer.
Check the actual target account for remaining quotas and VIP checkout. Record consumption per successful and failed job, quota refunds, export limits, expiration, and auto-renewal. For company or client work, verify whether a personal VIP entitlement permits organizational use. Paying for VIP does not grant the face, voice, script, music, font, template, and every output right by itself.
Pros
- Direct connection to Bilibili accounts and creator workflows.
- A browser process combines avatar, voice, script, and talking-video generation.
- Free quota can support a limited pilot before broader production.
- Account and identity gates provide more accountability than an anonymous cloning service.
- Repetitive, low-risk scripts for an owned personal brand can require less filming.
Cons
- Login is mandatory, and the logged-out page does not expose complete features, pricing, or quota.
- Free and VIP allowances are dynamic, making old reviews unreliable for budgeting.
- Identity, face, and voice data create concentrated biometric risk that cannot be reset like a password.
- Product-specific retention, training, verified deletion, model portability, and output rights remain unclear publicly.
- Lip sync, expression, pauses, and names can fail and require review of every output.
- Synthetic avatars can imply real presence or endorsement unless disclosure is prominent.
- A personal VIP plan, permission to publish on Bilibili, and enterprise commercial rights are separate questions.
Alternatives
| Tool | Best for | Advantage | Limitation |
|---|---|---|---|
| HeyGen | Multilingual marketing and avatar-video teams | Mature templates, translation, and commercial workflows | Overseas service; cost and biometric terms need review |
| Synthesia | Enterprise training and standardized explainers | More mature team controls and stock-avatar workflow | Custom likeness access and subscriptions can be expensive |
| Tavus | Developers building real-time video agents | Stronger real-time conversation and API integration | More engineering and strict controls for likeness and high-risk answers |
| D-ID | Quickly animating a talking image | Lightweight entry and broad API use | Consent, authenticity disclosure, and quotas still require governance |
FAQ
Is Bijian Studio still operating?
Yes. The official entry was reachable on July 21, 2026, but complete workspace and entitlement information requires a Bilibili account.
Is Bijian Studio completely free?
That is not a reliable long-term claim. Free and VIP accounts have dynamic allowances, and the current account page is the source of truth for generation count, duration, and features.
Does it require identity verification?
It is tied to the Bilibili account system, and functions involving a real face or voice may require identity checks. Follow the current logged-in flow and do not borrow accounts or bypass verification.
Can I clone another person’s face or voice?
Not without explicit, informed authorization. Public availability of a video does not constitute permission to create a model, synthesize new statements, or advertise with that identity.
How long are face and voice materials retained?
The logged-out public page does not adequately explain product-specific retention, derived features, or model deletion. Confirm those points in the current product agreement or a written support response before upload.
Can generated videos be used commercially?
Do not infer commercial rights from VIP access or a successful export. Rights are needed for the represented person’s face and voice, scripts and other assets, plus the product’s current model, template, and output terms.
Should synthetic videos be labeled?
Yes, where required by applicable law and platform policy, and whenever viewers could mistake the output for a real live appearance or personal endorsement. Some high-risk uses remain inappropriate even with a label.
Bottom Line
Bijian Studio is no longer accurately described as a waitlisted free beta. It is an active, account-gated Bilibili web avatar product with dynamic free and VIP quotas. It can reduce filming for low-risk content from an owned, consenting personal brand, but identity, face, voice, derived models, and outputs form a sensitive data chain. Obtain explicit consent, confirm product-specific retention, training, deletion, and rights, review every render, and disclose synthesis. If the current agreement or support channel cannot answer those questions, do not upload real biometric source material.