# Fix list: Alibaba Wan (Model Studio) From Anchor Terminal's listing at https://www.anchorterminal.com/tools/alibaba-wan, the October 2026 research run, assessed 1 October 2026. Grade B, 65.4 out of 100. This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public. For a coding agent working on Alibaba Wan (Model Studio): work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published. ## 1. Payments & pricing, 20 out of 100, up to 10 more on the total Why it scored 20: No x402, MPP or L402 (0). Per-second prices for every Wan model are published per region without a login, $0.05 to $0.20 a second for wan3.0-video in Singapore (20). A free quota exists in Singapore for 90 days, but the international sign-up asks for payment details first (0). A person signs up in a browser and creates the key (0). The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments): The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/). - 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which. - 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login. - 20, a free tier or trial that doesn't need a card. - 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API). Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied. Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol. ## 2. Reliability, 70 out of 100, up to 6 more on the total Why it scored 70: Alibaba Cloud runs a status page at status.alibabacloud.com with per-product components, which a third-party monitor reads as 2,736 components (20). The page renders in the browser only and we couldn't read its incident history. The monitor shows no Model Studio incident from July to September 2026, but that isn't the vendor's own record (5). The wan3.0-video and wan3.0-video-prime model pages list 300 requests a minute in every region, and Alibaba's Qwen Cloud page adds 5 concurrent tasks and a 500-task queue for wan3.0-video (15). The rate-limit guide documents 429 for bursts and says limits usually recover within a minute, with no Retry-After header, and the image-to-video guide says failed calls aren't billed (10). The Model Studio inference SLA promises 99.9 per cent monthly uptime with service credits on paid inference (10). Wan 3.0 carries no preview label in the docs (10). The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability): Hosted APIs, MCP servers, models and platforms. - 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own). - 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so. - 15, rate limits documented with numbers. - 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved. - 10, an SLA published for any paid tier. - 10, the surface agents use is generally available, not beta or preview. Local packages, SDKs, frameworks and stdio MCP servers. - 20, installs from an official package with supported runtimes stated. - 25, a public CI and test suite, passing on the default branch. - 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered). - 15, semver discipline and breaking changes called out in a changelog. - 15, version 1.0 or later, or declared stable. Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors. ## 3. Agent ergonomics, 63 out of 100, up to 6 more on the total Why it scored 63: Create and query calls return small JSON objects with a task id and a video URL (15). No webhook or callback is documented, so an agent polls, every 15 seconds per the docs (0 of the 10 for callbacks). Resolution, ratio and duration control output size (10). We found no task list endpoint in the video docs (0 of 10). The error page gives a code and a fix an agent can act on, though 429 guidance stops at retry later (18). No idempotency key, and the docs say not to resubmit but to keep polling. Failed calls aren't billed, so a retry after a failure is safe (10). Every parameter except the prompt or media has a default (5). The official Python SDK `dashscope` is current. We didn't confirm a second official SDK for video in this run (5). The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics): - 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries). - 20, pagination, filtering and output-size controls. - 20, actionable, documented error responses, codes and messages an agent can recover from. - 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations. - 15, sensible defaults, few required parameters, and official SDKs in at least two languages. Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs. ## 4. Schema & documentation, 67 out of 100, up to 5.4 more on the total Why it scored 67: No OpenAPI or other machine-readable contract for the DashScope video endpoints (0). llms.txt at docs.modelstudio.console.alibabacloud.com links Markdown copies of the Wan guides and references (10). The Wan 3.0 reference says which model to pick and which inputs combine, and older pages point readers to newer models, but there's little on when not to use a mode (14). Typed parameters with enums for resolution and ratio, ranges for duration (2 to 30) and seed, a 20,000 character prompt cap and stated defaults. `input` takes either a prompt or a media array, which a schema can't express (13). Six curl examples on the Wan 3.0 reference and an error-code page with HTTP status, code, message and a fix for each entry, including video checks such as InvalidFile.FPS (15). Dated snapshots such as `wan2.7-t2v-2026-06-12` and a dated model release page (15). The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema): APIs and MCP servers. - 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool). - 10, llms.txt or Markdown docs served for agents. - 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference. - 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs. - 0 to 15, examples and documented error responses. - 15, versioning and a public changelog. Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference. ## 5. Security & auth, 80 out of 100, up to 3.5 more on the total Why it scored 80: Model reading of the checklist (credential, training, retention, operator visibility, programme). API keys belong to a workspace, can be limited to up to 30 models and an IP allow-list of 20 entries, can be reset, and temporary keys last at most 1,800 seconds (30). The Model Studio privacy notice says Alibaba Cloud will never use customer data for model training (20). Result URLs and task ids expire after 24 hours (5). We found no deletion call and no retention period for prompts and inputs (0). Audit logs record every model call by default (request id, model, latency, status) and monitoring filters by API key (15). SOC 2 named in the Model Studio privacy notice (5) and a compliance repository linked from it (5). No security.txt on alibabacloud.com, and we didn't find a bug bounty page for Model Studio in this run (0). The checklist (https://www.anchorterminal.com/benchmark/#checklist-security): - 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option. - 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions. - 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10. - 0 to 15, audit logs or per-call visibility for the operator. - 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public. Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing. ## 6. Maintenance & community, 77 out of 100, up to 2 more on the total Why it scored 77: Model reading. wan3.0-video-prime arrived on 20 August 2026 and the `dashscope` SDK shipped 1.27.7 on 24 September, five releases in September (30). A published decommissioning policy gives 3 months' notice for mainline models and 30 days for dated snapshots, by email and announcement. The snapshot window is short, so 15 of 20. Dated model release page updated to 28 September, plus Alibaba Cloud ticket support (12 of 15). Current official SDK on PyPI (15). PyPI lists Python 3.9 to 3.14 but links no source repository or CI (5). The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance): - 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older. - 20, at least three releases or dated changelog entries in the last 90 days. - 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15. - 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models). - 10, package health, current dependencies and CI. Models are read for deprecation notice periods and model churn rather than release counts. ## 7. Transparency & trust, 80 out of 100, up to 1.8 more on the total Made of editorial 70, provenance 90. Why it scored 80: The API models are closed under Alibaba Cloud's product terms. Wan 2.1 and 2.2 weights are Apache 2.0 on GitHub, so part of the family can be self-hosted (20). The privacy notice and FAQ both say customer data is never used for training (10). Result retention is stated (24 hours), input retention isn't (5). The notice, FAQ and API reference agree, but we didn't find a DPA in this run (5). A decommissioning policy with notice periods by model type (20). Six regional hosts are documented with separate keys and prices. We found no subprocessor list (10). The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency): - 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms. - 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors). - 0 to 20, a deprecation policy or notices with dates. - 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted). The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two. Provenance checks not met in full (half of this category, computed from checked facts): - security.txt: not found (0 of 10) ## What we couldn't check What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it. - We couldn't read the incident history on status.alibabacloud.com, which renders in the browser only - The listing's free quota figures (30 seconds for Wan 3.0, 50 per Wan 2.x model) weren't confirmed. The free-quota page names no video amounts - Whether the Model Studio SLA explicitly covers video generation. It covers paid model inference and doesn't name models - Whether a Java SDK covers Wan 3.0, and whether the DashScope SDK has a blocking wait helper for video ## Weaknesses - Result URLs and task ids expire after 24 hours - No webhook or callback, so every job is polled - No OpenAPI document for the video endpoints - Keys, hosts and prices are per region, and Wan 3.0 uses a per-workspace host - Sign-up asks for payment details before the free quota ## What costs an agent a turn today The notes we give agents before they call it. Each one is a workaround an agent shouldn't need. - Send `X-DashScope-Async: enable` on the create call or it fails, then poll GET /api/v1/tasks/{task_id} about every 15 seconds - Copy the video out within 24 hours. The URL and the task id both expire - Use the host from your workspace and region. A Singapore key won't work on a Beijing host - Don't resubmit a slow job. Poll it. Failed jobs aren't billed, so resubmit only after FAILED - Set `audio` false if you don't want sound. It's on by default for Wan 3.0 ## What the review panel asked for - Callback URL - Single global endpoint - publish one global price table ## When it's done Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.