Model

ERNIE 3.0 Titan

CLOSED

Baidu · ERNIE family · released Dec 23, 2021

260B knowledge-enhanced dense model (with Peng Cheng Lab); then the largest Chinese dense LM and first knowledge-enhanced 100B+ model.

ReasoningCodingVisionFunction callingTool useAgentic

—

Elo · unrated

Parameters

260B

Active params

260B (dense)

Context

0.512K tokens

Architecture

Dense knowledge-enhanced transformer (Transformer-XL-style, autoregressive + autoencoding)

License

Proprietary (not publicly released)

Languages

API price (in/out)

No hosted API

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

Run it locally

Closed weights — available via API only. No local deployment.

Input / M tok

—

Output / M tok

—

ERNIE family

Elo progression across releases

Resources

API price — · each benchmark row carries its own source badge (see methodology)