AI provider

Baidu

Baidu is a publicly listed technology company headquartered in Beijing. Robin Yanhong Li and Eric Yong Xu founded the business in 2000. Its work spans internet search, AI cloud services and AI applications. Robin Li is chairman and chief executive officer.

Models & tools

  • Pretrained sparse language model with 280 billion total parameters and 14 billion active per token.

    Launch date Downloadable
  • Unlimited-OCR-2

    Baidu · Unlimited-OCR

    Document parsing model with sparse-reference sliding-window attention for long documents.

    Launch date Downloadable · API
  • Unlimited-OCR

    Baidu · Unlimited-OCR

    DeepSeek-OCR-based model for joint parsing of multiple pages in a single document-level pass.

    Launch date Downloadable
  • PaddleOCR-VL-1.6

    Baidu · PaddleOCR-VL

    A 0.9-billion-parameter document vision-language model with improved training for difficult document regions.

    Launch date Downloadable
  • ERNIE 5.1

    Baidu · ERNIE 5

    Hosted general-purpose model derived from ERNIE 5.0 through elastic model selection and task-specific post-training.

    Launch date API · Provider product
  • ERNIE 4.5 0.3B

    Baidu · ERNIE 4.5

    Post-trained dense language model with 300 million parameters for local text processing.

    Launch date not confirmedDownloadable
  • Dense 300-million-parameter base checkpoint in the ERNIE 4.5 language family.

    Launch date not confirmedDownloadable
  • Post-trained sparse language model with 21 billion total parameters and about three billion active per token.

    Launch date not confirmedDownloadable
  • Sparse language base model with a 21-billion-parameter expert pool and about three billion active per token.

    Launch date not confirmedDownloadable
  • ERNIE 4.5 21B A3B Thinking

    Baidu · ERNIE 4.5 Thinking

    Reasoning-trained 21B-total, 3B-active language model for complex problems and function calls.

    Launch date not confirmedDownloadable
  • Post-trained sparse text model with 300 billion total parameters and 47 billion active parameters.

    Launch date not confirmedDownloadable
  • ERNIE 4.5 base language model with 300 billion total parameters and 47 billion active parameters.

    Launch date not confirmedDownloadable
  • ERNIE 4.5 Turbo 128K

    Baidu · ERNIE 4.5 Turbo

    Hosted ERNIE 4.5 Turbo text model with a 128K context for long documents.

    Launch date not confirmedAPI
  • ERNIE 4.5 Turbo 32K

    Baidu · ERNIE 4.5 Turbo

    Hosted ERNIE 4.5 Turbo text-generation model with a 32K context.

    Launch date not confirmedAPI
  • ERNIE 4.5 Turbo VL

    Baidu · ERNIE 4.5 Turbo

    Hosted vision-language model with image understanding and a 128K context.

    Launch date not confirmedAPI
  • Post-trained sparse vision-language model with 28 billion total parameters and three billion active parameters.

    Launch date not confirmedDownloadable
  • Sparse vision-language base checkpoint with 28 billion total parameters and three billion active parameters.

    Launch date not confirmedDownloadable
  • Visual reasoning model with grounding, video understanding and image tools, trained from the 28B-A3B architecture.

    Launch date not confirmedDownloadable
  • Post-trained vision-language model with 424 billion total parameters and 47 billion active parameters.

    Launch date not confirmedDownloadable
  • Large sparse vision-language base model with 424 billion total parameters and 47 billion active parameters.

    Launch date not confirmedDownloadable
  • ERNIE 5.0

    Baidu · ERNIE 5

    Multimodal mixture-of-experts foundation model with 2.4 trillion parameters and a shared text, image, video and audio architecture.

    Launch date not confirmedAPI
  • Hosted ERNIE 5.0 preview with reasoning enabled by default and an optional non-thinking mode.

    Launch date not confirmedPreviewAPI
  • Hosted reasoning preview trained with an iterative mixed reinforcement-learning framework.

    Launch date not confirmedPreviewAPI
  • ERNIE-Image

    Baidu · ERNIE-Image

    Eight-billion-parameter diffusion image generator with a lightweight prompt enhancement component.

    Launch date not confirmedDownloadable
  • ERNIE-Image-Turbo

    Baidu · ERNIE-Image

    Distilled eight-billion-parameter image model that generates images in eight sampling steps.

    Launch date not confirmedDownloadable
  • NAVA

    Baidu · NAVA

    Joint audio and video generator with 6.3 billion parameters, built from Wan2.2-TI2V-5B.

    Launch date not confirmedDownloadable
  • PaddleOCR-VL-1.5

    Baidu · PaddleOCR-VL

    A 0.9-billion-parameter document model for skewed, warped or unevenly lit pages, with Tibetan and Bengali support.

    Launch date not confirmedDownloadable
  • Qianfan-OCR

    Baidu · Qianfan-OCR

    Four-billion-parameter document model for recognition, layout analysis, question answering and field extraction in 192 languages.

    Launch date not confirmedDownloadable
  • Qianfan-VL-3B

    Baidu · Qianfan-VL

    Vision-language model combining Qwen2.5-3B with an InternViT encoder for images up to 4K.

    Launch date not confirmedDownloadable
  • Qianfan-VL-70B

    Baidu · Qianfan-VL

    Seventy-billion-parameter Qianfan vision-language model built from Llama 3.1 and an InternViT image encoder.

    Launch date not confirmedDownloadable
  • Qianfan-VL-8B

    Baidu · Qianfan-VL

    Eight-billion-parameter vision-language model combining Llama 3.1 and InternViT with a 32K context.

    Launch date not confirmedDownloadable