ERNIE 5.5 Flash Preview scores 1,668 on WebDev

// 2h agoBENCHMARK RESULT

ERNIE 5.5 Flash Preview scores 1,668 on WebDev

Baidu’s 280B-parameter mixture-of-experts model, with 14B active parameters and a 1M-token context window, scored 1,668 on Code Arena’s WebDev leaderboard. It is available through Qianfan, while its promised Apache 2.0 weights have not yet been released.

// ANALYSIS

The result is a strong signal for Baidu in agentic coding, but calling ERNIE open-source is premature while the model remains API-only.

–Its 1M-token context and 14B active parameters target long-horizon coding with comparatively efficient inference.
–The 1,668 score narrowly exceeds DeepSeek-V4.1-Flash and Kimi K3 in the cited leaderboard snapshot.
–The ranking is provisional, using AutoEval with limited public voting data, so it should not be treated as definitive model quality.
–Qianfan access gives developers an immediate way to test coding, tool-use, and multi-step workflows.
–Apache 2.0 weights would materially change its competitive position by enabling local deployment and independent evaluation.

// TAGS

ernie-5.5-flash-previewllmbenchmarkai-codinglong-contextopen-weights

DISCOVERED

2h ago

2026-11-12

PUBLISHED

3h ago

2026-11-12

RELEVANCE

9/ 10

AUTHOR

hamzalaabar

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE37m ago

Framer Agent Generates Images, Shaders

Framer’s AI Agent can now turn natural-language prompts into generated imagery and interactive shader effects, including pointer-driven smudges, halftones, CRT visuals, bokeh, and pencil styles. The workflow aims to produce polished web interactions without manual code or After Effects.

OPEN SOURCE44m ago

Perplexity Decider v1.3 Adds Locate Extraction

Perplexity open-sourced its latest decision models under Apache licensing, adding Locate for extracting exact text spans and character positions. Both models are free on OpenRouter for one week and support classification, routing, moderation, reranking, and structured evaluation.

LAUNCH49m ago

Daybreak Red Team automates continuous pen tests

OpenAI’s Daybreak Red Team puts coordinated AI agents to work testing web applications, validating vulnerabilities, and checking whether patches actually stop exploits. The early-access service targets organizations that need continuous offensive security coverage without maintaining a dedicated red team.