Ling-3.0-flash: benchmark results

InclusionAI's open 124B MoE (5.1B active) with native hybrid-linear attention and hybrid reasoning, built for high-throughput agent workloads (July 2026). Provider: InclusionAI. Released 2026-07-23. Access: Open.

Unified ELO 1764 ± 14, rank #206 of 2650 rated models, from 152 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
ArtifactsBench77AVG (self-reported)100
LLM Stats (Multi-IF)87.7Score (%)100
Multi-IF87.7Average (self-reported)96.4
AA-Omniscience Hallucination Rate44.05Hallucination Rate (%)85.4
BenchmarkList ECI132.17Capability Index (ECI)80.8
Nejumi 4 - GLP - Information Retrieval80.45Score (%)80.5
Artificial Analysis Intelligence Index24.94Intelligence Index80.2
LLM Stats Score36.36LLM Stats Score (conservative rating)79.1
Nejumi 4 - GLP - Abstract Reasoning71Score (%)78.9
AA GPQA Diamond85.45Accuracy (%)78.7
Tinybird AI SQL Benchmark - First-Attempt Success Rate98Questions answered with a valid query on the first attempt (78.1
AA Humanity's Last Exam23.68Accuracy (%)75.4

Interactive version: theaggregate.ai/model?slug=ling-3-0-flash · How It Works · Data refreshed daily, snapshot 2026-09-20.