Ethical Stack.

Baidu Qianfan Model API

Qianfan offers API-first access to ERNIE and other hosted models, with OpenAI-compatible REST APIs and developer tooling.

Score calculation

(1 + 4 + 2 + 1 + 7) ÷ 5 = 3.0

5 of six dimensions rated. Unrated dimensions are excluded from the average.

Full scoring method ↗
Setup details

Baidu Qianfan model-service API, including ERNIE and other models available in the account; consumer Wenxin use is excluded.

What fits your needs

Baidu Qianfan model-service API, including ERNIE and other models available in the account; consumer Wenxin use is excluded.

Fit

Route applications to ERNIE or other available hosted models through a single API platform with OpenAI-compatible REST access.

See the evidence
Tradeoffs to expect

Tradeoff

The API is managed inference; available model names, limits, and account-level service rules can change.

See the evidence
How to switch

Getting started

Create a Qianfan account and API credentials, then follow the current V2 endpoint and selected model identifier in Baidu’s docs.

See the evidence
Practical questions

Data handling

Baidu’s cloud terms describe manual review of some model-call snippets for service improvement, subject to the stated permission context.

See the evidence

Energy practices

1/10↑

This grades published energy practices, not the service’s actual footprint.

See the evidence

The reviewed materials define the Qianfan hosted model API and its functions but provide no service-specific reporting on electricity, emissions, water, sourcing, or resource-reduction controls. Missing measurements don’t show high consumption.

1 + 0 reporting + 0 electricity + 0 controls = 1/10. This grades documented practices and transparency.

No comparable service-level electricity, emissions, water, location mix, or independently reviewed footprint is established in these sources.

Reporting0/3
No service-specific operating boundary or footprint disclosure appears in the reviewed product material.
Electricity sourcing0/3
No route-specific clean-electricity sourcing or hourly power evidence is established.
Resource controls0/3
No route-specific measured efficiency or documented resource-reduction control is established.

Water disclosure

Water reporting has not been assessed for this choice.

Training data

—Depends on the model↑

The hosted route does not identify one fixed model and training data.

See the evidence

This Qianfan hosted model API can expose multiple models or change its model selection over time. The product page is not a source-specific training-data and permissions record for every model served through the route.

Select the exact model and version, then review its training sources, permissions, creator controls, and any API-specific data terms.

Your data & privacy

4/10↑

Published terms describe service data and possible manual review for some model-improvement activity.

See the evidence

Baidu AI Cloud’s terms describe processing that can include human review of selected big-model conversation snippets for improving services, with permission language for that example. The terms also describe account, usage, and support records; this is meaningful disclosure, though it is broad cloud-platform language rather than a Qianfan-model retention schedule.

Published processing and human-review disclosures support a middle-low band; the general terms don’t confirm a narrow model-specific retention limit.

The applicable model-specific retention period, review rate, opt-out controls, and any region or account-plan exceptions are not resolved by these general platform terms.

Openness

2/10↑

This is a hosted product route; its interface documentation does not disclose the hosted model implementation or grant access to its weights.

See the evidence

The reviewed Qianfan hosted model API materials describe a hosted service, without publishing the service implementation or checkpoint weights for this route. Its openness rating concerns access to this product surface, not any separately released model family.

A hosted interface is documented, but source code, training artifacts, and route-specific downloadable weights are not established in these sources.

The exact model checkpoint and any route-specific release artifacts may change; this rating covers the named service route.

Worker protections

1/10↑

No model-specific worker safeguards are established for Baidu in the reviewed model sources.

See the evidence

The model release pages describe weights and technical properties, not conditions for employees, data annotators, evaluators, moderators, or infrastructure contractors. This is a source-bounded disclosure gap, not evidence of mistreatment. These findings concern the developer or operator. This product’s use in a particular deployment needs separate evidence.

1 base + 0 conditions + 0 worker voice/remedy + 0 evidence/follow-through = 1/10; this is a disclosure finding, not proof of abuse.

Worker locations, pay and conditions, supplier coverage, voice mechanisms, labor audits, and remediation outcomes are not established by these model-release sources.

Pay & conditions0/3
No worker-level conditions or supply-chain safeguards are established in this model-family source set. This is a disclosure gap, not proof of mistreatment.
Worker voice & remedy0/3
The model sources do not document a worker voice or remedy mechanism for data, evaluation, or infrastructure labor.
Evidence & follow-through0/3
No independent worker-level audit or remediation outcome is established in the reviewed model sources.
Workers rubric ↗

Military ties

7/10Not established↑

No route-specific military relationship or end-use oversight was established for this Baidu Qianfan model-service API, including ERNIE and other models available in the account; consumer Wenxin use is excluded..

See the evidence

The reviewed Baidu Qianfan Model API materials describe user access, features, or account rules but don’t confirm a military customer, military deployment, or a model-specific restriction on military use. This route finding does not resolve the wider company’s government work.

7 base + 0 safeguard adjustment = 7/10 (0 of 9 safeguard points; every 3 adds 1). No military relationship was established in the cited sources. This ranks above known defense ties without claiming a verified clean record.

A separate, current company-level review is needed for defense contracts; downstream deployments and any customer-specific safeguards are not established here.

Use restrictions0/4
The cited route terms don’t spell out a specific restriction on military or surveillance use.
Enforcement & oversight0/3
The cited route terms don’t show route-level enforcement of military end-use restrictions.
Relationship disclosure0/2
The cited route material does not disclose a military customer or deployment for this route.

The cited Baidu Qianfan Model API documentation covers product/API terms, but does not identify a military customer, deployment, or end-use restriction.

Military rubric ↗

BDS targets

Not listed↑

Baidu is not named in the two official BNC campaign lists checked. This is a limited list check, not BDS approval.

See the evidence

The official BNC consumer boycott/pressure guide and No Tech campaign page do not name Baidu among their current targets. BDS uses selective campaigns, so absence from these lists is neutral and should not be read as a clean record or approval. These findings concern the developer or operator. This product’s use in a particular deployment needs separate evidence.

This check covers only named targets in the two cited BNC campaigns; it does not cover all BDS campaigns, local groups, investments, suppliers, or downstream hosts. Lists and ownership may change.

More from Baidu

All sources

Side by side