inclusionAI Ling-2.6-flash | OpenRouter | Page 1

orchid grail Apr 21, 2026, 6:58 PM

#

https://openrouter.ai/inclusionai/ling-2.6-flash:free

Ling-2.6-flash (free) - API Pricing & Providers

Ling-2.6-flash is an instant (instruct) model from inclusionAI with 104B total parameters and 7.4B active parameters, designed for real-world agents that require fast responses, strong execution, and high token efficiency. $0 per million input tokens, $0 per million output tokens. 262,144 token context window, maximum output of 32,768 tokens.

rancid prairie Apr 21, 2026, 8:28 PM

#

For its to-be price (identical to Step 3.5 Flash) point, it gets mogged by nearly any other model. It may be good for agentic workloads, as it seems very optimized for that, and its speed is remarkable compared to other models in its price range. While it’s free if you have an agentic workload, it could be worth your time, but once the free period ends, I suspect this model will not be particularly compelling

spark ravine Apr 22, 2026, 7:46 PM

#

aaaand it's dying in a week thumbsupcat

rancid prairie Apr 22, 2026, 8:06 PM

#

Well that was fast

fossil crane Apr 22, 2026, 8:16 PM

#

underwhelming. for a "fast" model replies take 10 seconds direct on openrouter chat

low cloud Apr 23, 2026, 6:21 AM

#

after testing some Chinese sensitive topic, Ling's political censorship is same as:

GLM 5.1
Kimi K2.5
MiMo-V2.5-Pro

#

which is very weak

#inclusionAI Ling-2.6-flash