ModelBeat
Models/Ling-1T
All models
A

Ant Group: Ling-1T

Watch this modelCompareAnnouncement
ling-1t

Ling-1T is Ant Group’s open-weights AI model, released Oct 10, 2025.

Context
—
Input
— / 1M
Output
— / 1M
Released
Oct 10, 2025
Modalities

Benchmarks

Scores on standardized evaluations. Higher is better — and rank shows where Ling-1T lands among all models tracked on Model Beat.

12.0
Intelligence Index
Artificial Analysis
12th percentile of tracked models
14.0
Coding Index
Artificial Analysis
14th percentile of tracked models
8.0
Agentic Index
Artificial Analysis
8th percentile of tracked models

Reasoning

3 evals
GPQA Diamond71.9%

Graduate-level scientific reasoning

Humanity's Last Exam7.2%

Frontier of human expert knowledge

MMLU-Pro82.2%

Broad expert knowledge

Coding

2 evals
LiveCodeBench67.7%

Contamination-free coding

SciCode35.2%

Scientific research coding

Math

1 evals
AIME 2024/202571.3%

Olympiad-qualifier math

Agentic & Tools

1 evals
τ²-bench32.7%

Tool-agent-user reliability

Compare Ling-1T with

Ling-1T vs GPT-5.6 SolLing-1T vs Claude Fable 5Ling-1T vs Gemini 3.1 ProLing-1T vs Muse SparkLing-1T vs Grok 4.5

Frequently asked questions

What is Ling-1T?

Ling-1T is an AI model, from Ant Group, released Oct 10, 2025.

Who created Ling-1T?

Ling-1T is an AI model developed by Ant Group, released Oct 10, 2025.

How does Ling-1T perform on benchmarks?

On standardized evaluations tracked by Epoch AI, Ling-1T scores 82.2% on MMLU-Pro, 67.7% on LiveCodeBench, 35.2% on SciCode.

Is Ling-1T open source?

Ling-1T is released with open weights (Open weights (unrestricted)), so the model can be downloaded and self-hosted.

Benchmark data from Epoch AI (CC BY) and Artificial Analysis; pricing, specs & descriptions from OpenRouter. Data updated Jul 30, 2026.