TickerTain
$SPCX····$NVDA····$MSFT····$GOOGL····$ANTHROPIC·$TSLA····$CRWV····$NBIS····$HIMS····$TSM····$ASML····$FIGURE·$LYFT····$UBER····$SNOW····$META····$ELEVENLABS·$CURSOR·$COGNITION·$OPENAI·$AAPL····$RKLB····$XAI·$AMZN····$AMD····$000660.KS····$INTC····
← newsroom
$MANAS · China Just Blocked Meta. Now The Founders Can't Leave.
now playing · $MANAS
$MANASneutral· lowunknown
Chinese AI agent startup acquired by Meta for $2B, now blocked by Beijing with founders detained
Manas built the first consumer AI agent (2M waitlist in 7 days, $1.5K invite resales), moved to Singapore, raised $70M from Benchmark, hit $90M revenue run rate; China now claims…
$META···bearish· lowunknown
China blocks Meta's $2B Manas acquisition, stranding founders and disrupting agentic strategy
China's unprecedented unwinding of a closed deal signals AI assets are now treated as national security assets, leaving Meta's agentic layer and advertising revenue gains (30% qua…
$SCALE-AIneutral· lowunknown
Meta pays $15B for half of Scale AI to acquire Alexander Wang after other agentic pursuits fail
After failing to acquire Ilya Sutskever's SSI, Mira Murati's Thinking Machines, and Perplexity, Meta spent $15B for half of Scale AI primarily to secure Alexander Wang's expertise…
$DEEPSEEKbullish· lowunknown
DeepSeek V4 matches GPT-5.5 benchmarks, leads OpenRouter token usage as Chinese open-source models surge
DeepSeek's latest model achieves parity with top US closed models, and Chinese models now surpass US models in public token consumption on OpenRouter, suggesting developers prefer…
$HUAWEIbullish· lowunknown
Huawei chips catch up to Nvidia for inference workloads as China mandates domestic silicon for frontier models
Huawei's inference-optimized architecture now powers China's latest frontier models (DeepSeek V4, Kimi K2.6, Qwen 3.6), which match 90-95% of US model capability, validating China…
$NVDA···neutral· lowunknown
China mandates domestic inference chips, threatening Nvidia's dominance as open-source models drive GPU demand
China's top labs (DeepSeek, Moonshot, Alibaba) are mandated to train on Huawei chips optimized for inference, not pre-training, closing the hardware gap; meanwhile open-source mod…