ERNIE 5.0
ERNIE 5.0 is Baidu Wenxin's flagship native-multimodal model spanning text, image, video, and audio with strong reasoning and agent skills. Access it through Velokey's unified endpoint with below-list pricing.
Use cases
Coding & agent automation · Multimodal understanding & creation
Input
Text, image
Output
Text
Billing
Per token
Start a conversation
Type a message below to begin
Sign up to get $0.5 in free credits — about 20 free images
Pricing details
Transparent usage-based pricing with no hidden fees.
Detailed Pricing
Tiered pricingUsage-based pricing breakdown
| Spec | Price |
|---|---|
| Input/1M tokens | 0.9 Credits≈$0.9 |
| Output/1M tokens | 3.5 Credits≈$3.5 |
| Spec | Price |
|---|---|
| Input/1M tokens | 1.5 Credits≈$1.5 |
| Output/1M tokens | 5.8 Credits≈$5.8 |
Billing Rules
- Output price varies by model, resolution, quality, duration, or token usage.
- Uploaded reference files may be charged separately when supported.
- Add-on capabilities such as web search or image search may be billed per request.
- Token-based models are billed by input and output tokens.
- If the primary route is unavailable, Velokey may automatically switch to a stable fallback route when possible.
- Usage is pay-as-you-go with transparent credit deduction.
* Final cost depends on the generated result.
Related models
Explore other models in the same family.
BaiduSave 20%ERNIE 5.1
ERNIE 5.1 is an efficient sub-network of ERNIE 5.0 with much lower inference cost, strong agentic, reasoning, world-knowledge, and creative-writing abilities.
OpenAISave 20%GPT 5.4 Mini
OpenAI's fast, cost-efficient GPT-5.4 Mini for high-volume production, with vision, long context, and strong tool use at low latency.
AlibabaSave 20%Qwen3.6 Plus
Alibaba's Qwen3.6 Plus, a balanced flagship-tier model with strong reasoning, long context, multimodal input, and agentic tool use.
AnthropicSave 20%Claude Fable 5
Anthropic's Mythos-class frontier model, state-of-the-art across coding, agentic work, vision, and scientific research.
Affordable ERNIE 5.0 API
Access Baidu Wenxin's ERNIE 5.0 on Velokey — a native-multimodal, trillion-scale flagship with unified text, image, video, and audio, efficient sparse activation, and strong reasoning and agents, through a unified endpoint with below-list pricing.
Meet ERNIE 5.0 and its API access
ERNIE 5.0 is Baidu Wenxin's flagship model released on February 6, 2026. It is built on a native-multimodal architecture: a 2.4-trillion-parameter unified autoregressive network that handles text, image, video, and audio in one framework rather than fusing separate models. A Modality-Agnostic Routing MoE activates only about 3% of parameters per token, and an Elastic Training strategy allows variable depth, width, and sparsity for deployment flexibility. Velokey provides an affordable, easy-to-integrate ERNIE 5.0 API: a unified chat endpoint, a clear request flow, and below-list pricing for production systems, agent pipelines, and multimodal applications. On this platform you can send text, image, and video inputs and receive text output.
Native unified multimodal architecture
One unified autoregressive network handles text, image, video, and audio together — Next-Token Prediction for text, Next-Frame-and-Scale Prediction for images, and Next-Codec Prediction for audio — instead of stitching separate models together after the fact.
View Documentation →
Trillion-scale MoE efficiency
2.4 trillion parameters with a Modality-Agnostic Routing MoE that activates only about 3% of parameters per token, plus an Elastic Training strategy for variable depth, width, and sparsity — trillion-scale capacity that stays efficient to deploy.
View Documentation →
High-fidelity multimodal generation
Generates high-fidelity images, industry-leading video, and top-tier audio from one model — best-in-class on TUT2017 audio understanding with competitive text-to-speech, unifying understanding and creation.
View Documentation →
Knowledge, reasoning, coding & agents
Strong knowledge retrieval, logical reasoning, code generation, instruction following, and agent tool-calling, with a context window progressively extended from 8K to 128K tokens for long-horizon work.
View Documentation →
How to deploy the ERNIE 5.0 API on Velokey
Get started with just a few steps.
Sign up and get an API Key
Sign in to the Velokey console to generate an API Key that authenticates every request sent to ERNIE 5.0.
Configure request parameters
Set model to ernie-5-0 and submit messages. Pass text, image, and video inputs and receive text output, with context up to 128K tokens.
Integrate and start calling
Send requests to Velokey's unified /v1/chat/completions endpoint to power assistants, agents, and multimodal apps, with usage managed in one console.
Real-world use cases for ERNIE 5.0
From engineering to creative and analytical work, covering many high-value scenarios.
Coding & development assistants
Code generation and instruction following make it fit for development assistants, refactoring, and code review helpers.
Agents & tool-calling
Reliable agent tool-calling makes it fit for multi-step automation, orchestration, and long-horizon agent workflows.
Multimodal understanding & creation
Understands text, image, and video and generates high-fidelity images and industry-leading video — fit for content and creative studios.
Knowledge retrieval & Q&A
Strong knowledge retrieval fits research helpers, enterprise Q&A, and retrieval-augmented assistants over long documents.
Analysis & reasoning
Logical reasoning over a 128K context fits data analysis, structured extraction, and multi-document synthesis.
Audio & speech applications
Best-in-class TUT2017 audio understanding with competitive text-to-speech fits transcription, voice, and audio analysis flows.
Why choose Velokey for ERNIE 5.0
Discount on list price
Call ERNIE 5.0 at a lower price on Velokey — pay-as-you-go with no forced subscription, reducing upfront cost.
Unified API interface
One API Key accesses ERNIE 5.0 and other models — no juggling multiple accounts and keys, simplifying integration.
Complete docs and guidance
Endpoint references, parameter details, and multimodal examples help you integrate text, image, and video inputs smoothly.
Stable and highly available
Automatic failover and stable fallback routes maintain availability when the primary route is down, keeping production reliable.
API Reference
Complete API call examples and parameter descriptions
Endpoint
https://api.velokey.ai/v1/chat/completionsAuthentication
Bearer YOUR_API_KEYRequest Example
curl https://api.velokey.ai/v1/chat/completions \
-X POST \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ernie-5.0",
"messages": [{"role": "user", "content": "Hello!"}],
"temperature": 0.7
}'Response Example
{
"id": "req_abc123",
"model": "ernie-5.0",
"created": 1234567890,
"data": { ... }
}ERNIE 5.0 API FAQ
What makes ERNIE 5.0 a native-multimodal model?
It uses one unified autoregressive framework across modalities instead of fusing separate models afterward: Next-Token Prediction for text, Next-Frame-and-Scale Prediction for images, and Next-Codec Prediction for audio, all in a single network.
How large is the model and how is it efficient?
It has 2.4 trillion parameters, but a Modality-Agnostic Routing MoE activates only about 3% of them per token. An Elastic Training strategy supports variable depth, width, and sparsity for deployment flexibility, so trillion-scale capacity stays efficient to serve.
Which inputs and outputs are supported on Velokey?
On this platform you can send text, image, and video inputs and receive text output. The model itself also supports high-fidelity image, industry-leading video, and audio generation, and it is best-in-class on TUT2017 audio understanding with competitive text-to-speech.
How large is the context window?
The context window has been progressively extended from 8K to 128K tokens, so you can reason across long documents, extended conversations, and multi-step agent workflows.
How is it billed on Velokey?
This page's pricing area and your dashboard show the current Velokey price, billed by input and output tokens with a discount on list price. One API key and a unified endpoint keep integration simple.
Start building with ERNIE 5.0 today
Use one API key to access low-cost, stable AI models through Velokey.