FlashNEWS BRIEFING

Baidu Unveils ERNIE 5.5 Flash Preview; Weights to Be Open-Sourced by Nov 25

1 min read

88Z AI, 12 November — Baidu has released its first post-trained model built on ERNIE 5.5 Flash Base: 280B total parameters, 14B active, a 1M-token context window, with weights to be open-sourced by November 25.

At a glance

  • ERNIE 5.5 Flash Preview is a 280B-total / 14B-active MoE model with a 1M-token context and over 200 tokens/s generation on Qianfan.
  • Weights will be open-sourced under Apache 2.0 by November 25; developers can watch for download and evaluation.
  • It scores 1668 on Code Arena: WebDev, trending as the #2 open model.
Baidu's ERNIE 5.5 Flash Preview announcement
Baidu's ERNIE 5.5 Flash Preview announcement

Release Details

Baidu's ERNIE team has introduced ERNIE 5.5 Flash Preview, the first post-trained model built on the previously released ERNIE 5.5 Flash Base. It has 280B total parameters with 14B active per token, a 1M-token context window, and generation speed above 200 tokens/s per user on Baidu Qianfan.

According to @Baidu_Inc, the model uses a new architecture designed for efficient inference, delivering stronger performance in coding, 3D modeling, knowledge work, and content creation. It posts competitive scores on Terminal-Bench, AutomationBench, and other benchmarks. Weights will be released by November 25.

Community Response

The Arena platform shows ERNIE 5.5 Flash Preview trending toward becoming the #2 open model on the Code Arena: WebDev leaderboard at 1668 points, just above DeepSeek-V4.1-Flash (Max) at 1664 and Kimi K3 (Max) at 1658, and roughly on par with Qwen3.8-Max at 1670. The weights will be open-sourced under Apache 2.0.

Sources & editorial note

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.

  1. @Baidu_ERNIESource published

Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.

Related reading

All stories
88Z.ai