Qwen3.8-27B

# Fix list: Qwen3.8-27B

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/qwen3-8-27b, the October 2026 research run, assessed 20 October 2026. Grade B, 63.7 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 Qwen3.8-27B: 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. Security & auth, 45 out of 100, up to 9.6 more on the total

Why it scored 45: Read for open weights. File safety 30, safetensors only and no code from the repository runs on load. Integrity 15, files in the Qwen organisation with a SHA-256 for each shard and CRC32 sums for the small files. Safety evaluation 0, none published on the card. Guidance for agent use 0, none found. Security programme 0, no disclosure policy or security.txt on qwen.ai, and Alibaba's group-wide security response wasn't checked for this model.

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.

## 2. Maintenance & community, 25 out of 100, up to 6.6 more on the total

Why it scored 25: Read for open weights. Last change 14 August 2026, 67 days before this check (20). Changes that say what they fixed 0, none in the last 90 days. Responsiveness 0, none of the 40 most recent discussions that aren't pull requests had a reply from a Qwen member. Runtime support it maintains 0, the recipes are vLLM's and SGLang's and we found no first-hand record of Alibaba's own runtime code for this model. File health 5 of 10, open reports about the chat template and structured output (discussions 173 and 175) and an open pull request to change the template (203) have no reply.

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.

## 3. Schema & documentation, 63 out of 100, up to 6 more on the total

Why it scored 63: Read for open weights. Card 13 of 25, it covers intended use, architecture, parameters and context but not limitations, languages or a knowledge cut-off (4 off for each). A chat_template.jinja in the repository (20). Tool calling 10 of 20, the format is defined only by the template, and the parser to use is named by vLLM's and SGLang's recipes rather than by the card. Examples 10 of 15, request examples for text, image, video and non-thinking mode, serving commands on the linked recipes, and no tool-calling example. Thinking, reasoning effort and preserved thinking documented (10). Versions 0, the commit messages are uploads and README updates.

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.

## 4. Payments & pricing, 60 out of 100, up to 5 more on the total

Why it scored 60: Read for open weights, by the self-hosted rule, so 60 is the most a model can score here. No payment protocol (0). Apache 2.0, the full terms public with no login (20). Commercial use free (20). The files download without an account or a gate (20).

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.

## 5. Transparency & trust, 47 out of 100, up to 4.6 more on the total

Made of editorial 45, provenance 49.

Why it scored 47: Editorial half, read for open weights. Licence 30, Apache 2.0 with no acceptable use policy added. Training data 0, the card describes none. Evaluations 5 of 20, numbers in the card's tables with no set-up stated. Version policy 10 of 20, earlier Qwen models stay in their own repositories, with no stated policy or dated notes.

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):

- Domain age: qwen.ai, registered 2023-08-27 (3 years) (7 of 15)
- Status page: not found (0 of 10)
- Changelog: not found (0 of 10)
- security.txt: not found (0 of 10)

## 6. Agent ergonomics, 85 out of 100, up to 2.4 more on the total

Why it scored 85: Read for open weights. Context 262,144 tokens native (25). Tool calling it was built for, and the template renders several calls in one turn (25). Structured output 0, not documented for the model. Reasoning control, thinking off per request and reasoning effort low to xhigh (15). Fit 20, Ollama's default build is 18 GB and Alibaba's FP8 repository 30.9 GB.

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.

## 7. Reliability, 90 out of 100, up to 2 more on the total

Why it scored 90: Read for open weights. Runtimes 30, model code in vLLM, SGLang and mlx-lm, the qwen35 architecture in llama.cpp and a local qwen3.8 entry in Ollama's library (6 each). A reference implementation in a released Transformers, qwen3_5 in 5.20.0 (20). Sampling parameters for thinking and non-thinking modes, with serving recipes linked (15). Hardware 5 of 15, the card states no memory needs, so only the file sizes say it. Alibaba's own FP8 build (10). A general release (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.

## 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.

- unchecked: whether Alibaba's group-wide security response centre takes reports about Qwen models; no link from the card or qwen.ai was found
- unchecked: who wrote the Transformers and vLLM support; we found no first-hand record that Alibaba maintains it
- The hosted Qwen Cloud version is described on the card as coming soon and isn't listed
- Performance and Task success are pending until our local runs; every build fits our 128 GB machine

## Weaknesses

- The card doesn't describe tool calling, so the parser to use comes from vLLM's and SGLang's recipes
- No training data, limitations, languages, knowledge cut-off or safety results on the card
- No memory needs stated by Alibaba
- None of the 40 most recent discussions had a reply from a Qwen member, including reports about the chat template
- The files and card haven't changed since 14 August 2026

## 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.

- Serve it with `--tool-call-parser qwen3_coder` and `--reasoning-parser qwen3`; the Hermes parser leaves its tool calls unparsed, per SGLang's cookbook
- Set `enable_thinking: false` in `chat_template_kwargs` for short tasks, since thinking is on by default
- Use the sampling the card gives for each mode, `temperature=1.0, top_p=0.95` thinking and `temperature=0.7, top_p=0.8, presence_penalty=1.5` without
- Turn on YaRN only for contexts beyond 262,144 tokens; the card says static YaRN can cost quality on short texts
- Check JSON output yourself, since structured output isn't documented for the model

## What the review panel asked for

- answer the template reports
- dated release notes
- memory figures on the card
- a tool-calling section
- a knowledge cut-off
- safety results on the card
- a way to report a flaw

## 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.