Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Huawei / Pangu / Pangu NLP

Pangu-NLP-N2-128K-5.0.3.2

Versioned Pangu language asset for managed training and text inference.

Language models (LLMs)
availableClosed weightsText generation
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
8 of 18 specification fields collected
Context window128KSee specification notes below
Maximum outputNot collectedAwaiting source review
Input billingSee billing notesHuawei Cloud ModelArts Studio
Output billingSee billing notesAccess and billing notes below

Specifications

Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.

Context / input limit
128KSource for Context / input limit (opens in a new tab)

Provider sequence-length notation retained; input and output share this budget.

Maximum output
Not collected
Input modalities
TextSource for Input modalities (opens in a new tab)
Output modalities
TextSource for Output modalities (opens in a new tab)
Reasoning
SupportedSource for Reasoning (opens in a new tab)
Tool calling
Not collected
Structured output
Not collected
Streaming
SupportedSource for Streaming (opens in a new tab)
Batch processing
Not collected
Fine-tuning
Not supportedSource for Fine-tuning (opens in a new tab)
Knowledge cutoff
Not collected
Release date
Not collected
Parameters
Not collected
Architecture
Not collected
Languages
Not collected
Context caching
Not collected
Published price
Provisioned inference resources; numeric rate not verifiedSource for Published price (opens in a new tab)

Inference resources use monthly/yearly unit subscriptions. Applicable model subscriptions and training/data resources are separate; an exempt model-subscription fee does not mean free inference.

Documented tool behavior
Provider reports improved tool capability in this releaseSource for Documented tool behavior (opens in a new tab)

No exact function-calling request contract or per-tool limits established by this statement.

Access & pricing

Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.

Huawei Cloud ModelArts Studio

Provider source (opens in a new tab)
API model ID
Not collected

Asset names identify catalogue versions. Callable model names and deployment IDs are assigned by the deployment; the asset name is not substituted for an API selector.

Access
Huawei Cloud account, applicable asset subscription and provisioned inference resources required. Some subscriptions require public-beta approval; account eligibility must be checked in the console.
Regions
China, Southwest Guiyang 1 (catalogue region); not a general residency guarantee.
assetName
Pangu-NLP-N2-128K-5.0.3.2Source for assetName (opens in a new tab)
apiPaths
POST /v1/{project_id}/deployments/{deployment_id}/chat/completions, POST /api/v2/chat/completionsSource for apiPaths (opens in a new tab)
API model selector
V2 uses Deployed_Model from the deployed service details; V1 uses the deployment IDSource for API model selector (opens in a new tab)
outputBudget
Input plus generated output must fit the model context; no independent fixed output cap specifiedSource for outputBudget (opens in a new tab)
streamRestrictions
Incremental output with stream=true; n must be 1Source for streamRestrictions (opens in a new tab)
Authentication
V1: X-Auth-Token or X-Apig-AppCode; V2: Bearer API keySource for Authentication (opens in a new tab)
continuedPretraining
Not supportedSource for continuedPretraining (opens in a new tab)
modelCompression
SupportedSource for modelCompression (opens in a new tab)
minimumInferenceUnits
4Source for minimumInferenceUnits (opens in a new tab)

Model deployment requirement, not a universal resource-order minimum or price.

edgeDeployment
Not supportedSource for edgeDeployment (opens in a new tab)

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Retention, training use, regional processing and enterprise terms have not been summarized for this endpoint.

Caching

Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.

01 / INITIAL INPUTProcess & retain

The system processes eligible input and retains reusable state.

02 / LATER REQUESTCheck available state

Reuse depends on matching input, session state and the cache policy.

03 / CACHE HITReuse & generate

Reusing state can reduce repeated work. Any charges follow the billing rules below.

Huawei Cloud ModelArts Studio

Not verified
How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Not collected
Limits & data boundaries
Not collected

No verified cache rate for this endpoint. An empty rate is not a zero charge.

What the measurements say

No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.

Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.

The record behind the model

Why it is included

Inferred closed-weight classification for this exact named managed Pangu asset in the developer catalogue, reviewed against Huawei’s separately published OpenPangu releases. The reviewed release does not provide matching public weights. A shared Pangu brand, model-training service or downloadable customer adaptation does not establish public availability of this original asset.

Classification evidence (opens in a new tab)

Sources reviewed