Ling 3.1 Flash (Reasoning)

by InclusionAI

Ling 3.1 Flash (Reasoning) is a reasoning model from InclusionAI, released on September 30, 2026. It ranks #12 of 220 models by Intelligence Score with a score of 51.

Ling 3.1 Flash (Ling-3.1-flash) is InclusionAI / Ant Group's hybrid-reasoning MoE flash model: ~560B total parameters with ~25B active per token. API-only at launch (weights not yet on Hugging Face). Hosted context is commonly 262K (256K-class); 1M context is the announced design target, not what all gateways serve today. Text→text; hybrid reasoning / tool-using agent positioning. Vendor launch benches (self-reported): GDPval-AA v2.1 1673 Elo, SkillsBench 68.7, AutomationBench 52.5, Terminal-Bench 4.0 40.4, CyberGym 87.9, FrontierSWE 75.16, SWE Atlas Codebase QnA 55.92, Finance Agent v2 57.87, HealthBench Professional 65.35, DRACO 85.49, MultiChallenge 69.78. No stable paid price row in an existing first-party/gateway provider dir in this repo (Vercel AI Gateway / Novita free trial noted but not stored) — available_in_zeroeval false.

Key specs

ProviderInclusionAI
Release date2026-09-30
Licenseopen weights
Intelligence Score51
Coding index38.8

Quick answers

Is Ling 3.1 Flash (Reasoning) open source?

Yes. Ling 3.1 Flash (Reasoning) is an open model released under the open weights license.

When was Ling 3.1 Flash (Reasoning) released?

InclusionAI released Ling 3.1 Flash (Reasoning) on September 30, 2026.

How does Ling 3.1 Flash (Reasoning) compare to Claude Opus 5.5 (Reasoning)?

Ling 3.1 Flash (Reasoning) scores 51 on the Intelligence Score versus 60.5 for Claude Opus 5.5 (Reasoning), the current #1.

Compare on the AI leaderboard · All AI models

More models from InclusionAI