Realtime AI News
DeepSeek V4 Flash Tops 7 Trillion Weekly Calls, Ranks No.1 Globally
DeepSeek V4 Flash has surpassed 7 trillion calls per week, ranking first globally in weekly call volume, according to a 36Kr report. The milestone positions the model as the most frequently used large language model in real-world production today.
DeepSeek V4 Flash has logged more than 7 trillion calls per week, ranking first globally in weekly call volume, according to a report from Chinese tech media 36Kr. The report, headlined "Over 7 Trillion Calls Weekly: DeepSeek V4 Flash Ranks No.1 Globally in Weekly Call Volume," points to the model's sheer scale of real-world usage.
Weekly call volume is one of the most direct measures of genuine adoption for a large language model. Unlike benchmark scores or demo videos, it reflects how frequently developers and end users actually invoke the model in production, and higher numbers mean more real business scenarios have been connected to it.
DeepSeek V4 Flash is DeepSeek's model offering, and topping the weekly call ranking makes it one of the most frequently invoked large models in the world. For a model positioned around fast response, trillion-level weekly traffic also suggests it has held up under high-frequency, low-latency workloads.
The signal matters for the industry: when a model logs trillions of weekly calls, it is running steadily inside real applications rather than living only in benchmarks and demos. Leadership in call volume often says more about a model's standing in the developer ecosystem than a single performance edge.
The 36Kr report provides corroboration from a major Chinese tech outlet, though the item circulates through a Google News aggregation link, so readers may want to cross-check the figure against DeepSeek's official channels and other usage statistics.
What to watch next is whether DeepSeek V4 Flash can sustain its lead in the coming weeks and whether the momentum pulls more developers into building on top of it. At the same time, competition over call volume among top models will continue to shape inference pricing and the broader ecosystem.
Why it matters
DeepSeek V4 Flash's No.1 ranking in weekly call volume signals massive real-world usage and could cement DeepSeek's position in the inference services market.
Nearby Updates
All08/05, 16:04
NVIDIA Nemotron 3 Ultra Highlights Major AI Agent Advances
Coverage from Techgenyz spotlights NVIDIA's Nemotron 3 Ultra as a major step forward in AI agent capabilities. The release positions agent-focused abilities as the headline advance in NVIDIA's high-end open model line, with benchmark results and technical specifics still pending.
08/05, 15:05
Sand.ai Open-Sources What It Calls the First 100B-Parameter MoE Video Generation Model
Sand.ai has open-sourced a Mixture-of-Experts video generation model it bills as the world's first 100-billion-parameter MoE video model, with 114B total parameters and just 6B active. The model generates 10-second 1080p clips at a reported cost of about 0.5 yuan each, sharply lowering the cost barrier for high-quality AI video.
08/05, 14:33
Microsoft Halts 'Tokenmaxxing' With Strict Budget Caps; GPT-5.6 Becomes Internal Default
Chinese tech outlet QbitAI reports that Microsoft has halted internal "Tokenmaxxing" — pushing token usage to its limits — and locked AI budgets, with over-limit usage now at employees' own risk. The report adds that GPT-5.6 has become Microsoft's default internal model.
08/05, 13:59
ByteDance Seed Unveils SeedRealtime: Full-Duplex Audio-Video Model Debuts in Doubao
ByteDance's Seed team has released SeedRealtime, a full-duplex audio-video large model now integrated into the Doubao app. The model lets users watch, listen, and speak simultaneously, removing the awkward pauses of turn-based voice assistants and pushing real-time multimodal interaction into a mainstream consumer product.