ERNIE 5.0 Thinking Preview

ERNIE 5.0 Thinking Preview

Baidu · Released Nov 13, 2025
Intelligence #629 / 756
20.3 our score
AA Index #181 / 438
14.3 Artificial Analysis
Input Price
Not priced
Output Price
Not priced
Context
Not reported

Analysis Summary

ERNIE 5.0 Thinking Preview from Baidu has the strongest measured capability in this batch, with excellent mathematics, coding, general knowledge, and tool-interaction results. Its terminal result is also useful, although instruction following and long-context retention are only moderate. The preview designation signals that production reliability still needs validation.

The model is suited to technical reasoning, code generation, structured tool workflows, and agent prototypes where teams can add testing and safeguards. It is less proven for long-document work and customer-facing content that requires consistent tone and precise instruction adherence. No pricing, speed, context-window, or deployment information is provided, so procurement decisions should wait for those details and a reliability trial.

Assessed September 7, 2026

Editorial notes

ERNIE 5.0 Thinking Preview shows strong mathematics, coding, tool interaction, and general knowledge results. Its preview status and limited long-context evidence make it better for controlled technical evaluation than immediate default deployment.

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?

DFO Verdict

ERNIE 5.0 Thinking Preview shows strong mathematics, coding, tool interaction, and general knowledge results. Its preview status and limited long-context evidence make it better for controlled technical evaluation than immediate default deployment.

#629 of 756 overall

Benchmark scores

GPQA Diamond 77.7%
HLE 12.7%
MMLU Pro 83%
AIME 2025 85%
SciCode 37.5%
LiveCodeBench 81.2%
TerminalBench Hard 25%
τ²-Bench 83.9%
IFBench 41.4%
LCR 6.7%

Magenta = intelligence · Ink = technical/agentic · Cyan = content & long-context · Grey = community benchmarks. Data: Artificial Analysis, Hugging Face.

14.3 Intelligence Index·85 Math Index

How ERNIE 5.0 Thinking Preview compares

ERNIE 5.0 Thinking Preview ranks #181 of 438 AI models we track for overall intelligence. ERNIE 5.0 Thinking Preview is currently free to use via OpenRouter.

Position in the field
Intelligence: smarter than 17% of models #629
worst in fieldmedianbest in field
Price vs frontier peers · $ per 1M tokens
ERNIE 5.0 Thinking Preview $0.00 in $0.00 out
OpenAI: GPT-6 Astra $10.00 in $50.00 out
Anthropic: Claude Fable 5.1 $10.00 in $50.00 out
Claude Opus 5 $5.00 in $25.00 out

Dark bar = input · light bar = output, scaled to the priciest peer.

Context window vs peers · tokens
ERNIE 5.0 Thinking Preview 0

1M tokens ≈ 8 full-length novels or ~2,500 pages of business documents in a single request.

Intelligence3.4Technical0Value0Content4.8
Performance profile

Strongest on business fit. The pulled-in value corner is the trade-off, and if the shape matters more than the price, this is your model.

Compare shapes side-by-side →

Frequently asked questions about ERNIE 5.0 Thinking Preview

How much does ERNIE 5.0 Thinking Preview cost?

ERNIE 5.0 Thinking Preview is currently available for free via OpenRouter.

Who created ERNIE 5.0 Thinking Preview?

ERNIE 5.0 Thinking Preview is developed by Baidu and was released on November 13, 2025.

© 2026 Design for Online Ltd. Registered in England and Wales No. 10328553. VAT Registered. Design for Online® and Forerunner® are registered trademarks.