Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Baidu / ERNIE

ERNIE 5.0

Text and visual reasoning service with a 128K context and a scheduled China endpoint retirement on October 29, 2026.

Language models (LLMs)
deprecatedClosed weightsText generationVisionReasoning
Official documentation (opens in a new tab)Reviewed 12 Oct 2026
10 of 19 specification fields collected
Input token limit119KSee specification notes below
Maximum output65.5K65,536 tokens
Input billingSee ratesBaidu Qianfan China
Output billingSee ratesPublished units and conditions 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)
Maximum output
65,536 tokensSource for Maximum output (opens in a new tab)
Input modalities
Text, ImageSource 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
SupportedSource for Tool calling (opens in a new tab)

Explicitly listed in the China tool-calling support range. Do not infer the international support matrix.

Structured output
Not collected
Streaming
SupportedSource for Streaming (opens in a new tab)

The listed models use the Chat Completions API with SSE.

Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
Not collected
Parameters
Not collected
Architecture
Not collected
Languages
Not collected
Context caching
Not collected
Maximum input tokens
119KSource for Maximum input tokens (opens in a new tab)
Maximum thinking tokens
60KSource for Maximum thinking tokens (opens in a new tab)

max_tokens covers reasoning plus final output; ERNIE thinking models do not support thinking_budget.

Scheduled retirement
China endpoint: 2026-10-29. Scheduled after this snapshot.Source for Scheduled retirement (opens in a new tab)

Access & pricing

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

Baidu Qianfan China

Provider source (opens in a new tab)
API model ID
ernie-5.0
Access
Qianfan account and appropriate API key required. Documented access has not been verified with a paid inference call.
Regions
China platform; a hostname alone does not establish data residency.
API endpoint URL
POST https://qianfan.baidubce.com/v2/chat/completionsSource for API endpoint URL (opens in a new tab)
Request limits
RPM = 60 TPM = 150000Source for Request limits (opens in a new tab)

Published default quotas, not a universal service guarantee.

Scheduled retirement
2026-10-29Source for Scheduled retirement (opens in a new tab)

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 12 Oct 2026
CategoryUsageRateUnit
Native API inferenceInput up to 32k1CN¥6.001 million tokens
Native API inferenceInput above 32k2CN¥10.001 million tokens
Native API inferenceOutput up to 32k1CN¥24.001 million tokens
Native API inferenceOutput above 32k2CN¥40.001 million tokens

Rate conditions

  1. Input at most 32k tokens. Published CNY per 1,000 tokens, normalized to per million without currency conversion. Input limits still apply; k notation is retained as published.
  2. Input above 32k and at most 128k tokens. Published CNY per 1,000 tokens, normalized to per million without currency conversion. Input limits still apply; k notation is retained as published.
Pricing source (opens in a new tab)

Data & enterprise terms

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

Baidu Qianfan International

Provider source (opens in a new tab)
API model ID
ernie-5.0
Access
Qianfan account and appropriate API key required. Documented access has not been verified with a paid inference call.
Regions
International platform; availability and residency depend on the account and contract.
API endpoint URL
POST https://api.baiduqianfan.ai/v1/chat/completionsSource for API endpoint URL (opens in a new tab)
Input modalities
TextSource for Input modalities (opens in a new tab)

The tiered rate comes from the text-generation section. Vision and deep-thinking sections quote different rates.

Endpoint caveat
Still listed in the international catalog. No international retirement date was verified; the model-level deprecated badge reflects the China notice.Source for Endpoint caveat (opens in a new tab)
INPUT / 1M TOKENSFrom $0.84
OUTPUT / 1M TOKENSFrom $3.38

Input at most 32k tokens. Published USD per million tokens. Input limits still apply; k notation is retained as published. Above 32k input, input/output rates are USD 1.4/5.6 per million. Exact integer k boundaries are not inferred. Cached input price: not collected.

USD · observed 12 Oct 2026 · Pricing evidence (opens in a new tab)
Published billing units · original currency · observed 12 Oct 2026
CategoryUsageRateUnit
Native API inferenceInput up to 32k1$0.841 million tokens
Native API inferenceInput above 32k2$1.401 million tokens
Native API inferenceOutput up to 32k1$3.381 million tokens
Native API inferenceOutput above 32k2$5.601 million tokens

Rate conditions

  1. Input at most 32k tokens. Published USD per million tokens. Input limits still apply; k notation is retained as published.
  2. Input above 32k and at most 128k tokens. Published USD per million tokens. Input limits still apply; k notation is retained as published.
Pricing source (opens in a new tab)

Data & enterprise terms

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

Baidu Qianfan International

Provider source (opens in a new tab)
API model ID
ernie-5.0
Access
Qianfan account and appropriate API key required. Documented access has not been verified with a paid inference call.
Regions
International platform; availability and residency depend on the account and contract.
API endpoint URL
POST https://api.baiduqianfan.ai/v1/chat/completionsSource for API endpoint URL (opens in a new tab)
Billing basis
Published vision rate: USD 1.40 input / 5.60 output per million. The regional text table has separate input-length tiers.Source for Billing basis (opens in a new tab)
INPUT / 1M TOKENS$1.40
OUTPUT / 1M TOKENS$5.60

Published USD per million tokens. Input limits still apply; k notation is retained as published. Cached input price: not collected.

USD · observed 12 Oct 2026 · Pricing evidence (opens in a new tab)
Published billing units · original currency · observed 12 Oct 2026
CategoryUsageRateUnit
Native API inferenceInput1$1.401 million tokens
Native API inferenceOutput1$5.601 million tokens

Rate conditions

  1. Published USD per million tokens. Input limits still apply; k notation is retained as published.
Pricing source (opens in a new tab)

Data & enterprise terms

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

Baidu Qianfan International

Provider source (opens in a new tab)
API model ID
ernie-5.0
Access
Qianfan account and appropriate API key required. Documented access has not been verified with a paid inference call.
Regions
International platform; availability and residency depend on the account and contract.
API endpoint URL
POST https://api.baiduqianfan.ai/v1/chat/completionsSource for API endpoint URL (opens in a new tab)
Billing basis
Published reasoning rate: USD 1.40 input / 5.60 output per million. The regional text table has separate input-length tiers.Source for Billing basis (opens in a new tab)
INPUT / 1M TOKENS$1.40
OUTPUT / 1M TOKENS$5.60

Published USD per million tokens. Input limits still apply; k notation is retained as published. Cached input price: not collected.

USD · observed 12 Oct 2026 · Pricing evidence (opens in a new tab)
Published billing units · original currency · observed 12 Oct 2026
CategoryUsageRateUnit
Native API inferenceInput1$1.401 million tokens
Native API inferenceOutput1$5.601 million tokens

Rate conditions

  1. Published USD per million tokens. Input limits still apply; k notation is retained as published.
Pricing source (opens in a new tab)

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.

Baidu Qianfan China

ernie-5.0
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
The reviewed material does not establish model-specific cache eligibility, TTL, minimum prefix or discounted rates for this record. Generic API usage fields do not prove support.Source for Limits & data boundaries (opens in a new tab)

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

Baidu Qianfan International

ernie-5.0
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
The reviewed material does not establish model-specific cache eligibility, TTL, minimum prefix or discounted rates for this record. Generic API usage fields do not prove support.Source for Limits & data boundaries (opens in a new tab)

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

Baidu Qianfan International

ernie-5.0
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
The reviewed material does not establish model-specific cache eligibility, TTL, minimum prefix or discounted rates for this record. Generic API usage fields do not prove support.Source for Limits & data boundaries (opens in a new tab)

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

Baidu Qianfan International

ernie-5.0
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
The reviewed material does not establish model-specific cache eligibility, TTL, minimum prefix or discounted rates for this record. Generic API usage fields do not prove support.Source for Limits & data boundaries (opens in a new tab)

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 for this exact named Baidu commercial model from native documentation and comparison with Baidu’s public model repositories. No weights for this exact service were found in the reviewed inventory. Public ERNIE 4.5 and ERNIE Image releases are excluded, and ambiguous Turbo mappings remain under review. This is not proof that all Baidu models are closed.

Classification evidence (opens in a new tab)

Lifecycle dates

Retirement announced (China)24 Sept 2026Source (opens in a new tab)
China endpoint retirement29 Oct 2026Source (opens in a new tab)

Sources reviewed