A Chinese AI developer that makes the chatbot and model "Kimi." It is a regular fixture at the top of open-source model rankings. Headquartered in Beijing, this startup is, despite its name "Moonshot," a software company with no connection to space exploration or aerospace business. Its flagship model Kimi has been released with an emphasis on its ability to handle long context in a single pass, and it is often grouped together with DeepSeek, Alibaba's Qwen, and others as part of the lineage of Chinese-made open-source models. As of August 2026, OpenAI and Anthropic have China AI
Current rank (1M)
36
Rank over the last 30 days
2026.08.20 – 2026.09.18 · High 21 · Low 36 · Now 36
OpenAI published a Cognition case study on September 11. Devin, the autonomous software engineer, now uses GPT-6 Astra to test the code it writes and hands back a simulator recording and a test report as evidence.
Fugu Max routes each task to the leanest model that can solve it and undercuts rival output pricing by 40 to 60 percent. Fugu Ultra v2 posted its best scores with Fable 5.1 and GPT-6 Astra kept out of its pool.
Cognition's SWE-2 takes China's Kimi K3 and runs it through another round of reinforcement learning. The company says it lands within a point of Fable 5.1 while costing 64% less.
Across 7.1 million verification images, a training-free discriminator matched or beat existing detectors
Tencent says the model beat GPT-5.6 on a coding evaluation and even helped optimize its own training and inference process
Apple unveils Mac Studio with M5 Max/Ultra and Mac mini with M6/M5 Pro, targeting local large models through clustering
Since the high-cost coding model Fable arrived, developers are increasingly routing different tasks to different models depending on cost and job type
'Transfer station' proxies that defeat even selfie verification are undermining regional blocks and identity checks
Install one download file on your laptop and your existing CLI agent turns into a clone that acts like you
Community tokenizer comparison shows 60/60 match with zero average error — but that doesn't confirm it's the same model
Even after silencing unspoken numbers, some models still transcribed them correctly 30-40% of the time
Per-token price runs 4.4x the competitor average, yet developers keep choosing Anthropic's models
