Baidu Models

Q: Which Baidu model should I start with?

ernie-4.5-0.3b at $0.0136/M input — the cheapest entry here that declares tool calling. Move up to ernie-image-turbo when answer quality matters more than cost, or to ernie-5.0-thinking-preview for long-form reasoning.

23 modelsGeneral models from $0.0136/M inputUp to 183K context

Usage

Last 30 days · 2026-08-09 to 2026-09-07

Tokens

223M

Requests

43.3K

Models in use

22 of 23

Tokens per day, stacked by model

qianfan-ocr
ernie-5.1
paddleocr-vl-0.9b
ernie-5.0
ernie-5.0-thinking-preview
ernie-x1.1-preview
pp-structurev3
ernie-4.5-turbo-vl
2 more models

Which models that traffic went to

Share of 223M tokens. 12 models with traffic report no token counts and cannot be ranked here, including ernie-4.5-0.3b and ernie-4.5 — they are in the request view.

The two views disagree on purpose: a model can take a large share of the calls and a small share of the tokens — many short requests — or the reverse. Which one matters depends on whether your cost is driven by call volume or by prompt length. Measured on AIHubMix over the last 30 days, counting the 23 model IDs listed on this page; traffic routed through upstream-specific IDs that are not in the public catalog is not included.

All 23 Baidu Models

Open in model list

Baidu models on AIHubMix with input and output modalities, context length, maximum output, price per million tokens including cache read and cache write rates, and measured throughput and latency.
	Modalities
ernie-5.0-thinking-preview	Takes text, returns text.	183K	64K	$0.822$3.288/M	$0.822/M	72 tok/s	6.53 s
ernie-4.5	Takes text, vision, returns text.	160K	64K	$0.068$0.272/M	—	99 tok/s	0.95 s
ernie-4.5-turbo-vl	Takes text, vision, returns text.	139K	16K	$0.4$1.2/M	—	6 tok/s	0.75 s
ernie-4.5-turbo-latest	Takes text, vision, returns text.	135K	12K	$0.11$0.44/M	—	18 tok/s	0.62 s
ERNIE-X1.1-Preview	Takes text, returns text.	119K	64K	$0.136$0.544/M	—	14 tok/s	2.31 s
ernie-5.1	Takes text, returns text.	119K	64K	$0.5634$2.5353/M	$0.5634/M	—	—
ernie-5.0	Takes text, vision, audio, video, returns text.	119K	64K	$0.8219$3.2877/M	$0.8219/M	—	—
ernie-5.0-thinking-exp	Takes text, vision, returns text.	119K	64K	$0.8219$3.2877/M	$0.8219/M	—	—
ernie-x1-turbo	Takes text, returns text.	51K	28K	$0.136$0.544/M	—	18 tok/s	9.98 s
qianfan-ocr	Takes text, vision, returns text.	32K	28K	$0.062$0.248/M	—	—	—
qianfan-ocr-fast	Takes text, vision, returns text.	32K	28K	$0.664$2.7383/M	—	—	—
paddleocr-vl-0.9b	Takes , returns text.	—	—	Free$0.025/M	—	—	—
pp-structurev3	Takes , returns text.	—	—	Free$0.025/M	—	—	—
ernie-4.5-0.3b	Takes text, vision, returns text.	—	—	$0.0136$0.0544/M	—	—	—
embedding-v1	Takes text. Output modality not published.	—	—	$0.068$0.068/M	—	—	—
tao-8k		—	—	$0.068$0.068/M	—	—	—
ernie-4.5-turbo-128k-preview	Takes text, vision, returns text.	—	—	$0.108$0.432/M	—	—	—
ernie-x1.1-preview		—	—	$0.136$0.544/M	—	14 tok/s	2.31 s
qianfan-qi-vl		—	—	$0.2$0.6/M	—	—	—
baidu/ERNIE-4.5-300B-A47B	Takes text, vision, returns text.	—	—	$0.32$1.28/M	—	—	—
cc-ernie-4.5-300b-a47b	Takes text, vision, returns text.	—	—	$0.32$1.28/M	—	—	—
ernie-image-turbo	Takes , returns vision.	—	—	$2$2/M	—	—	—
musesteamer-air-image	Takes , returns vision.	—	—	$2$2/M	—	—	—

Prices are USD per million tokens; cache read and cache write are the rates for prompt-cache hits and for writing a prompt into the cache. Throughput and latency are measured on AIHubMix — the same figures the model detail page shows — not vendor claims. A dash means the catalog does not publish that field for that model, which is not the same as the model not supporting it.

Common Comparisons

ERNIE-X1.1-Preview vs baidu/ERNIE-4.5-300B-A47BERNIE-X1.1-Preview vs cc-ernie-4.5-300b-a47bERNIE-X1.1-Preview vs ernie-4.5ERNIE-X1.1-Preview vs qwen3.8-max-previewERNIE-X1.1-Preview vs kimi-k3

Baidu on AIHubMix

Which Baidu model should I start with?

ernie-4.5-0.3b at $0.0136/M input — the cheapest entry here that declares tool calling. Move up to ernie-image-turbo when answer quality matters more than cost, or to ernie-5.0-thinking-preview for long-form reasoning.

Which of these models reason before answering?

6 of the 23 models here declare a reasoning phase — they work through the problem before producing an answer, which helps on multi-step problems at the cost of extra output tokens. Use the Reasoning filter above the table to see them. The catalog does not record anything further about how they differ, so this page does not sort them into families.

Why are there several entries for the same model?

Because each row is a route you can call, not a model release. Some IDs name an upstream (azure-, alicloud-, cc-), some are the open-weight repository form (baidu/…), and some differ only in capitalisation, kept so older integrations keep working.

The catalog does not carry a field saying which of those a given row is, so this page does not sort them into buckets it would have to invent. Every row shows that route’s own price, context and speed — compare those directly, and open a model to see the upstreams that serve it.

Do I need a separate Baidu account?

No. One AIHubMix key covers every model on this page, and switching between them is a change to the model string — billing, rate limits, and logs stay in one place.