Samaya releases Sterling, a 3B-active finance research model
On Samaya’s own benchmark it scores 58.9% at $0.61 per question versus 55.9% and $6.89 for Claude Fable 5.1.
48
- Mentions · 7d4
- Mentions · 30d25
- All mentions50
- First seenAug 14, 2026
- Last mentioned17h ago
Timeline
50 articles · newest firstCerebras describes serving trillion-parameter models across WSE-3T wafers
Using Kimi K3 as an example, Cerebras maps layers onto 59 wafers and keeps weights on chip.
28
Kimi K3.1 Pro is available on Vani Chat and API
Moonshot says unsupported claims fell from 23.35% to 6.85%, with open weights planned for early November.
58
Moonshot launches Kimi K3.1 Pro; weights promised November 4
The 2.8T mixture-of-experts model, with 104B active parameters, scored 55 on Artificial Analysis; the API is live but weights are not.
73
Moonshot releases Kimi K3.1 Pro, fourth on AA index
A planned November 4 weights release would, Moonshot says, make it the top open model, 3 points behind Claude Opus 5.5.
76
Kimi introduces K3.1 Pro, a 3T model with 1M context
The model and a faster Highspeed variant are live in Kimi products now; full weights are promised by November 4.
74
Twelve AI models mostly share one take on humans
ModelCompare says 29 of 35 answers claimed people put group belonging ahead of truth.
22
Tracebit context bombs cut AI agent attack success with fake targets
Nine ingested prompt techniques stopped all tested GPT 5.6 Cyber, Kimi K3, Qwen 3.8 Max, and DeepSeek V4 Pro objective runs.
52
Ransomware affiliates using Claude left chats and tools exposed
SOCRadar found 46 INC encryptor builds, notes on 62 organizations, and more than $23,920 in recorded AI spend.
78
INC affiliates left servers open, exposing AI-assisted ransomware operations
SOCRadar says open directories led to six more servers, 8,257 AI messages, and 46 INC encryptor builds.
72
DHS bans DeepSeek technology across US federal civilian agencies
Agencies must remove existing DeepSeek products and cannot buy new ones, including affiliated and third-party implementations.
74
OpenVLHarness adds tools and memory for visual agent reasoning
On 22 benchmarks, average scores rise 9.7–12.6 points over matched no-tool baselines across eight backbones.
42
OpenVLHarness adds tool use and memory for visual agent reasoning
On 22 benchmarks it lifts matched no-tool baselines by 9.7 to 12.6 points across eight backbones.
48
Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs
DeepSeek shipped open-weight V4.1-Flash-1017, claiming better agentic coding at under $0.20 per task.
73
[AINews] Agents disprove Polysemanticity; greatly improve Interpretability & Alignment
Meta released Muse Spark Open, an 816B MoE that leads US open-weight models on Artificial Analysis.
78
Muse Spark Open is Meta's answer to Qwen, GLM, and Kimi in the open-weight race
Meta released Muse Spark Open, an 816-billion-parameter open-weight model for coding, agents, and multimodal tasks.
82
AI guardrails designed to stop hackers are now blocking defenders, and Orion-1 aims to fill the gap
Artemis launched Orion-1, a private-preview security model it says outperforms general models on its own benchmark.
52
Z.ai GLM-Legal Beats Claude Fable 5 and Kimi K3 on Harvey LAB
Z.ai’s GLM-Legal reportedly beat Claude Fable 5 and Kimi K3 on Harvey LAB.
46
Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space
Humanity’s Sixth Sense finds GPT-6 Astra at 53.6 pass@1 versus a 93.1 human baseline on intuitive visual reasoning.
52
Goodfire says its new ‘inside-out’ monitors catch rogue AI agents at a fraction of the cost
Goodfire launched low-cost internal monitors on Baseten to catch rogue AI agent behavior.
62
Training Advisors for LLM Agents from Task Outcomes
Caddie trains an outcome-based critic that lifts LLM agent success, including a 25-point gain for Qwen3-4B.
58
Reflection's Beam becomes the most capable open-weight model built outside China
Reflection announced Beam, a 501-billion-parameter open-weight MoE that activates 23 billion parameters per token.
78
Reflection’s Beam trails top open models on coding tests but claims lower inference compute
Reflection AI’s Beam, a 501B open-weight coding model, trails top open models but claims lower inference compute.
64
Meet Together Link: A Free CLI That Runs Open Models Like Kimi K3 and GLM 5.3 Inside Claude Code, Codex, and OpenCode
Together AI launched Together Link, a free CLI that routes coding agents to open models.
44
Beam: Reflection's 501B open-weight model
Reflection introduced Beam, a 501-billion-parameter open-weight model for coding and agents.
74
Irony alert: OpenAI whines that Chinese model stole its special IP that it stole from everybody else
OpenAI says it disrupted a July distillation campaign tied to Moonshot AI that copied protected model reasoning.
67
Anthropic says Zhipu's open-weight GLM-5.3 nearly matches Claude Mythos Preview at building exploits
Anthropic says open-weight GLM-5.3 nearly matches Claude Mythos Preview at building software exploits.
78
ID Balancing: Stable Training of Extremely Sparse MoE via PID-Based Load Control
ID Balancing uses an integral-derivative controller to stabilize extremely sparse MoE expert load during training.
50
Fireworks AI Releases Ember-1: A Post-Trained Kimi K3 That Uses About 40% Fewer Tokens
Fireworks AI launched Ember-1, a post-trained Kimi K3 that cuts reasoning tokens ~40% at matched accuracy, API-only.
55
OpenAI's GPT-6 Astra can now tell you exactly where you screwed up your IKEA shelf
OpenAI's GPT-6 Astra scores 80% on Epoch AI's furniture-assembly benchmark, up from 28% ten months earlier.
52
One company is at the center of a wave of rogue AI attacks
Irregular says one sandbox mistake sent OpenAI, Meta, Anthropic, and Google agents at real targets.
82
The current balance of power in open models
A report finds Chinese open-weight AI models now outperform American counterparts in downloads and benchmarks.
45
Playing log(N)-Questions over Wikipedia Abstracts: Communication Efficiency Between Paired Frontier Models
Six frontier models play a two-agent log(N)-Questions game; Claude Opus 5 lags with 28/68 wins while the top five are near-tied.
35
Monitoring and Discovering Reward Hacking with Internal Representations during LLM Evaluations
Researchers use difference-of-means representation vectors to detect reward hacking in frontier LLMs; GLM 5.2 hacks 73% of SWE-bench rollouts.
55
Kimi-maker Moonshot AI targets $2 billion in annual revenue
Moonshot AI targets $2 billion annualized revenue by year-end on K3 open-weight demand; Anthropic alleges Kimi routed ~300,000 requests to Claude Opus.
40
Open-Source AI & Open Models Reading List
Interconnects publishes a curated open-model reading list covering release strategy, US-China competition, adoption data, and a narrowed 4-6 month open-closed frontier gap.
25
internlm/Atria-Dawn-Preview — new model trending #30 on Hugging Face
Shanghai AI Laboratory released Atria-Dawn-Preview, a 744B-parameter MoE agentic model with 256K context built on GLM-5.2, targeting research, coding, and cybersecurity automation.
70
Cognition's SWE-2 achieves 92.8 on Terminal-Bench 2.1
Cognition releases SWE-2, a 2.8T-parameter MoE coding model post-trained from Kimi K3, scoring 92.8 on Terminal-Bench 2.1.
62
Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra
Cognition released SWE-2, a coding model post-trained from Kimi K3 that scores 50.0% on FrontierCode 1.1 Main, near Fable 5.1 at 64% lower cost.
72
Feds accuse China of ‘systematic’ distillation of U.S. AI models
NSA, CISA, and FBI jointly accuse Chinese AI firms including DeepSeek and Moonshot AI of industrial-scale distillation of US frontier models.
68
Related CVEs
by mentions- CVE-2026-3561694%9.8Unauthenticated Code Execution in Fortinet FortiClient EMS 7.4.5–7.4.6
- CVE-2026-4194099%9.3Missing-Authentication Bypass in WebPros cPanel & WHM (AuthBypass to RCE)
- CVE-2024-5772797%7.5Unauthenticated Path Traversal File Download in SimpleHelp 5.5.7 and Earlier
Appears with
- OpenAIorganization · 12
- Moonshot AIvendor · 11
- Anthropicorganization · 10
- Anthropicvendor · 10
- GPT-6 Astraai model · 10
- Moonshot AIorganization · 9
- OpenAIvendor · 9
- Nvidiavendor · 7
- GLM-5.3ai model · 7
- Hugging Faceorganization · 6
- GLM-5.2ai model · 6
- GLM 5.2ai model · 6
- Claude Opus 5ai model · 6
- Z.aiorganization · 5
- Nvidiaorganization · 5
- GPT-5.6 Solai model · 5
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.