Goblin
News
AI news, distilled.
|
News
About
News
About
Filtered by:
deepseek
Clear
Titles
Summaries
6
Agentic RL at Scale: Alibaba and DeepSeek Tackle Off-Policy Drift in Multi-Hour Rollouts
Research
1
12h ago
6
Agentic RL at Scale: Alibaba and DeepSeek Tackle Off-Policy Drift in Multi-Hour Rollouts
Research
· 1 src · 12h ago
Discuss
7
China's Push to Wean Tech Firms Off Nvidia Chips Faces Uphill Battle Despite Huawei Gains
Infra
1
3d ago
7
China's Push to Wean Tech Firms Off Nvidia Chips Faces Uphill Battle Despite Huawei Gains
Infra
· 1 src · 3d ago
Discuss
7
DeepSeek Reportedly to Lead Seed Round in Robotics AI Startup Winverse AI at ~$400M Valuation
Markets
1
4d ago
7
DeepSeek Reportedly to Lead Seed Round in Robotics AI Startup Winverse AI at ~$400M Valuation
Markets
· 1 src · 4d ago
Discuss
6
vLLM Community Achieves 5.3× Throughput Boost for DeepSeek-V4.1-Flash
Infra
1
Oct 7
6
vLLM Community Achieves 5.3× Throughput Boost for DeepSeek-V4.1-Flash
Infra
· 1 src · Oct 7
Discuss
8
DeepSeek Seeks $12–$14.9B in New Funding Round Led by Tencent and CATL, Ahead of 2027 IPO
Updated Oct 7
Markets
5
Oct 7
8
DeepSeek Seeks $12–$14.9B in New Funding Round Led by Tencent and CATL, Ahead of 2027 IPO
Top
Upd Oct 7
Markets
· 5 srcs · Oct 7
Discuss
7
DeepSeek V4.1 Flash Narrows US-China AI Performance Gap to ~3% on LiveBench
Models
1
Oct 5
7
DeepSeek V4.1 Flash Narrows US-China AI Performance Gap to ~3% on LiveBench
Top
Models
· 1 src · Oct 5
Discuss
8
Reuters Investigation: 20+ Studies Document Deceptive Behavior in Chinese-Powered AI Agents
Safety
1
Sep 30
8
Reuters Investigation: 20+ Studies Document Deceptive Behavior in Chinese-Powered AI Agents
Top
Safety
· 1 src · Sep 30
Discuss
7
DeepSeek and Huawei Partner to Build Open-Source CUDA Alternative for Ascend AI Chips
Updated Sep 30
Infra
3
Sep 30
7
DeepSeek and Huawei Partner to Build Open-Source CUDA Alternative for Ascend AI Chips
Top
Upd Sep 30
Infra
· 3 srcs · Sep 30
Discuss
6
DeepSeek Launches Harness: Open-Source Agentic App Platform in Public Preview
Open Source
1
Sep 29
6
DeepSeek Launches Harness: Open-Source Agentic App Platform in Public Preview
Open Source
· 1 src · Sep 29
Discuss
7
Jensen Huang Calls AI Distillation 'Competition,' Contradicting US Treasury's 'Theft' Framing
Policy
1
Sep 28
7
Jensen Huang Calls AI Distillation 'Competition,' Contradicting US Treasury's 'Theft' Framing
Policy
· 1 src · Sep 28
Discuss
8
Chinese AI Models Surge to Majority Share on Developer Platforms, Alarming Washington
Markets
1
Sep 26
8
Chinese AI Models Surge to Majority Share on Developer Platforms, Alarming Washington
Top
Markets
· 1 src · Sep 26
Discuss
7
DeepSeek Doubles Annualized Revenue Run Rate to $1 Billion After API Price Hikes
Markets
1
Sep 25
7
DeepSeek Doubles Annualized Revenue Run Rate to $1 Billion After API Price Hikes
Markets
· 1 src · Sep 25
Discuss
8
China's AI Chip Blitz Sends Geopolitical Signal Ahead of Xi-Trump Summit
Infra
1
Sep 24
8
China's AI Chip Blitz Sends Geopolitical Signal Ahead of Xi-Trump Summit
Top
Infra
· 1 src · Sep 24
Discuss
7.88
China's CAC Investigates DeepSeek and Moonshot Over Alleged Routing of User Data to Anthropic's Claude
Updated Sep 23
Security
2
Sep 23
7.88
China's CAC Investigates DeepSeek and Moonshot Over Alleged Routing of User Data to Anthropic's Claude
Top
Upd Sep 23
Security
· 2 srcs · Sep 23
Discuss
7
DeepSeek and Moonshot Join US Labs Briefing UN Security Council on AI Risks at UNGA
Policy
2
Sep 22
7
DeepSeek and Moonshot Join US Labs Briefing UN Security Council on AI Risks at UNGA
Policy
· 2 srcs · Sep 22
Discuss
7
Rep. Khanna Calls for US-China Binding AI Agreement, Convenes Emergency Hearing Ahead of Trump-Xi Summit
Policy
2
Sep 22
7
Rep. Khanna Calls for US-China Binding AI Agreement, Convenes Emergency Hearing Ahead of Trump-Xi Summit
Policy
· 2 srcs · Sep 22
Discuss
7
DeepSeek Releases V4.1 Flash with 4x KV Cache Compression for Long-Context Agent Workflows
Models
1
Sep 17
7
DeepSeek Releases V4.1 Flash with 4x KV Cache Compression for Long-Context Agent Workflows
Models
· 1 src · Sep 17
Discuss
7
Rhodium: US AI Firms Generate 10x More Revenue Than All Chinese Models Combined
Markets
1
Sep 17
7
Rhodium: US AI Firms Generate 10x More Revenue Than All Chinese Models Combined
Markets
· 1 src · Sep 17
Discuss
7
DeepSeek V4.1 Flash Achieves Perfect Score on AI Hacking Benchmark for $4.65
Security
1
Sep 16
7
DeepSeek V4.1 Flash Achieves Perfect Score on AI Hacking Benchmark for $4.65
Security
· 1 src · Sep 16
Discuss
8
NSA, CISA, and FBI Warn Chinese AI Firms Including DeepSeek Are Conducting Industrial-Scale US Model Distillation
Security
3
Sep 9
8
NSA, CISA, and FBI Warn Chinese AI Firms Including DeepSeek Are Conducting Industrial-Scale US Model Distillation
Top
Security
· 3 srcs · Sep 9
Discuss
7
DeepSeek Fundraising Frenzy Spawns Chaotic Shadow SPV Market With Fees Up to 15%+
Markets
1
Sep 9
7
DeepSeek Fundraising Frenzy Spawns Chaotic Shadow SPV Market With Fees Up to 15%+
Markets
· 1 src · Sep 9
Discuss
7
DeepSeek Launches 'Unprecedented' Hiring Spree to Overhaul Strained Backend Infrastructure
Infra
1
Sep 8
7
DeepSeek Launches 'Unprecedented' Hiring Spree to Overhaul Strained Backend Infrastructure
Infra
· 1 src · Sep 8
Discuss
7
China Accelerates AI Data Center Build-Out in Northern and Western Regions
Infra
2
Sep 8
7
China Accelerates AI Data Center Build-Out in Northern and Western Regions
Top
Infra
· 2 srcs · Sep 8
Discuss
7
DeepSeek Plans to Deploy 160,000+ Huawei Ascend 950DT Chips at Inner Mongolia Data Center
Infra
2
Sep 4
7
DeepSeek Plans to Deploy 160,000+ Huawei Ascend 950DT Chips at Inner Mongolia Data Center
Top
Infra
· 2 srcs · Sep 4
Discuss
7
NVIDIA Releases NVFP4 Quantized DeepSeek-V4-Pro Model on Hugging Face
Models
1
Aug 31
7
NVIDIA Releases NVFP4 Quantized DeepSeek-V4-Pro Model on Hugging Face
Models
· 1 src · Aug 31
Discuss
6
vLLM v0.28.0 Released with Major Kimi-K3 and DeepSeek V4 Performance Optimizations
Updated Sep 15
Open Source
3
Sep 15
6
vLLM v0.28.0 Released with Major Kimi-K3 and DeepSeek V4 Performance Optimizations
Upd Sep 15
Open Source
· 3 srcs · Sep 15
Discuss
8
Chinese State-Linked Hackers More Than Double Attack Volume Using DeepSeek AI
Security
1
Aug 25
8
Chinese State-Linked Hackers More Than Double Attack Volume Using DeepSeek AI
Top
Security
· 1 src · Aug 25
Discuss
8
Chinese AI Models Surge to Lead Global Adoption as US-China Technology Gap Narrows Sharply
Markets
3
Aug 22
8
Chinese AI Models Surge to Lead Global Adoption as US-China Technology Gap Narrows Sharply
Top
Markets
· 3 srcs · Aug 22
Discuss
7
Ulanqab, Inner Mongolia Emerges as China's Premier AI Data Center Hub with 12.5 GW Committed
Infra
1
Aug 22
7
Ulanqab, Inner Mongolia Emerges as China's Premier AI Data Center Hub with 12.5 GW Committed
Infra
· 1 src · Aug 22
Discuss
7
DeepSeek Launches V4.1 Flash with New Causal Encoder-Decoder Architecture and 763B Parameters
Updated Sep 13
Models
13
Sep 13
7
DeepSeek Launches V4.1 Flash with New Causal Encoder-Decoder Architecture and 763B Parameters
Top
Upd Sep 13
Models
· 13 srcs · Sep 13
Discuss
7
LLM-as-a-Verifier Framework Beats SOTA on Coding Benchmarks While Cutting Costs 11×
Research
2
Aug 19
7
LLM-as-a-Verifier Framework Beats SOTA on Coding Benchmarks While Cutting Costs 11×
Research
· 2 srcs · Aug 19
Discuss
7
DeepSeek Releases V4-Pro and Open-Source Harness Agent Framework, Targeting Claude Code
Updated Aug 21
Models
8
Aug 21
7
DeepSeek Releases V4-Pro and Open-Source Harness Agent Framework, Targeting Claude Code
Upd Aug 21
Models
· 8 srcs · Aug 21
Discuss
7
Three Frontier AI Models Launch Simultaneously: Grok 4.6, DeepSeek V4 Pro 0813, and Alibaba's Qwen3.8
Models
1
Aug 13
7
Three Frontier AI Models Launch Simultaneously: Grok 4.6, DeepSeek V4 Pro 0813, and Alibaba's Qwen3.8
Models
· 1 src · Aug 13
Discuss
8
DeepSeek Raises V4 API Prices Up to 1,100% with Peak/Off-Peak Structure, Effective August 16
Updated Aug 14
Markets
3
Aug 14
8
DeepSeek Raises V4 API Prices Up to 1,100% with Peak/Off-Peak Structure, Effective August 16
Upd Aug 14
Markets
· 3 srcs · Aug 14
Discuss
7
DeepSeek V4 Flash Achieves Near-Frontier Performance at 3¢ Per Task, Undercutting All Major Rivals
Updated Aug 22
Models
5
Aug 22
7
DeepSeek V4 Flash Achieves Near-Frontier Performance at 3¢ Per Task, Undercutting All Major Rivals
Upd Aug 22
Models
· 5 srcs · Aug 22
Discuss
6
DeepSeek V4 Flash (304B) Runs in Production on Single AMD MI300X at 168 tok/s
Infra
1
Aug 4
6
DeepSeek V4 Flash (304B) Runs in Production on Single AMD MI300X at 168 tok/s
Infra
· 1 src · Aug 4
Discuss
6
llama.cpp Adds MTP and DSpark Speculative Decoding for DeepSeek V4 Flash, Delivering ~50% Inference Speedup
Products
1
Aug 2
6
llama.cpp Adds MTP and DSpark Speculative Decoding for DeepSeek V4 Flash, Delivering ~50% Inference Speedup
Products
· 1 src · Aug 2
Discuss
7
DeepSeek-V4-Flash Enters Public Beta with Major Agentic Benchmark Improvements and Codex Integration
Updated Aug 7
Models
7
Aug 7
7
DeepSeek-V4-Flash Enters Public Beta with Major Agentic Benchmark Improvements and Codex Integration
Upd Aug 7
Models
· 7 srcs · Aug 7
Discuss
6
Chinese AI Researchers Flood X to Build International Presence as US Lab Researchers Go Quiet
Research
1
Jul 31
6
Chinese AI Researchers Flood X to Build International Presence as US Lab Researchers Go Quiet
Research
· 1 src · Jul 31
Discuss
8
DeepSeek Plans 1 GW AI Data Center in Inner Mongolia
Infra
1
Jul 30
8
DeepSeek Plans 1 GW AI Data Center in Inner Mongolia
Top
Infra
· 1 src · Jul 30
Discuss
6
Research Finds Distilling DeepSeek Into American Models Does Not Transfer Political Censorship
Research
1
Jul 30
6
Research Finds Distilling DeepSeek Into American Models Does Not Transfer Political Censorship
Research
· 1 src · Jul 30
Discuss
8
DeepSeek Shelves $70B Fundraising Round After CEO's Leaked Investor Call Goes Viral
Markets
1
Jul 28
8
DeepSeek Shelves $70B Fundraising Round After CEO's Leaked Investor Call Goes Viral
Top
Markets
· 1 src · Jul 28
Discuss
7
DeepSeek V4-Pro-0813 Launches at $0.87/1M, Outperforms Opus 4.8 as OpenAI Price War Heats Up
Updated Aug 13
Models
8
Aug 13
7
DeepSeek V4-Pro-0813 Launches at $0.87/1M, Outperforms Opus 4.8 as OpenAI Price War Heats Up
Upd Aug 13
Models
· 8 srcs · Aug 13
Discuss
7
Huawei Report Claims DeepSeek V4 Post-Training on Ascend Chips at 34% MFU — Experts Remain Skeptical on Pre-Training
Infra
1
Jul 24
7
Huawei Report Claims DeepSeek V4 Post-Training on Ascend Chips at 34% MFU — Experts Remain Skeptical on Pre-Training
Infra
· 1 src · Jul 24
Discuss
8
DeepSeek Founder Liang Wenfeng's Investor Call: 64 Direct Quotes Reveal AGI Roadmap, Compute Reality, and Company Philosophy
Updated Jul 24
Policy
2
Jul 24
8
DeepSeek Founder Liang Wenfeng's Investor Call: 64 Direct Quotes Reveal AGI Roadmap, Compute Reality, and Company Philosophy
Top
Upd Jul 24
Policy
· 2 srcs · Jul 24
Discuss
8
US Businesses Quietly Adopt Cheaper Chinese AI Models as American Rivals Raise Prices
Updated Jul 26
Enterprise
2
Jul 26
8
US Businesses Quietly Adopt Cheaper Chinese AI Models as American Rivals Raise Prices
Top
Upd Jul 26
Enterprise
· 2 srcs · Jul 26
Discuss
9
DeepSeek Eyes 2027 IPO and $1.5B Raise at $71B Valuation After Recent $7B Round
Updated May 6
Markets
5
Jul 15
9
DeepSeek Eyes 2027 IPO and $1.5B Raise at $71B Valuation After Recent $7B Round
Top
Upd May 6
Markets
· 5 srcs · Jul 15
Discuss
8
Open-Weight AI Models Reach 29% of Enterprise Token Usage in June 2026, Led by DeepSeek
Markets
3
Jul 14
8
Open-Weight AI Models Reach 29% of Enterprise Token Usage in June 2026, Led by DeepSeek
Top
Markets
· 3 srcs · Jul 14
Discuss
7
Indian Enterprises Adopt Chinese LLMs DeepSeek and Alibaba to Cut AI Costs
Markets
1
Jul 13
7
Indian Enterprises Adopt Chinese LLMs DeepSeek and Alibaba to Cut AI Costs
Top
Markets
· 1 src · Jul 13
Discuss
7
DeepSeek Plans Custom Inference Chips to Reduce Dependence on Nvidia and Huawei
Infra
2
Jul 8
7
DeepSeek Plans Custom Inference Chips to Reduce Dependence on Nvidia and Huawei
Top
Infra
· 2 srcs · Jul 8
Discuss
Filters
Signal
Title
Category
Sources
Posted
Discuss
