Foundation Models
Free
Kimi K2
Models (8)
- Kimi K3chatElo 1473
- Kimi K2.6chatElo 1456
- Kimi K3.1 Prochat
- Kimi K2.7 Codecode
- Kimi K2.5chat
- Kimi K2 Thinkingreasoning
- Kimi K2 0905chat
- Kimi K2 0711chat
Why is this hot
16 mentions / 7d across 4 sources (diversity ×1.45)last 48h: 4 vs baseline 2.8/day
- did you guys see the kimi engineer cosplaying deepseekbluesky
- Sources: Moonshot AI cancelled a $150M contract with a Southeast Asian neocloudrss
- Show HN: TrnRT – An open-source LLM inference engine for AWS Trainiumhn
- Misalignment signals in Kimi K3: What probes flag and where they peakhn
- Meta FAIR Self-Improving Memory Hits 56.0 on Continual Learning Benchrss
Top mentions by weighted contribution (source authority × engagement × recency).
Description
Moonshot AI's open-weight frontier model.
Recent mentions in news (15)
- Show HN: NLP capabilities of Opus, Astra, Argon measured by Codenames skillshn-ai · 12d
- Ask HN: Can anyone run frontier AI models at usable speeds?hn-ai · 13d
- The LLENS Platform for On-Premises Large Language Model Agents in a Japanese University Hospital: Development and Evaluation Studyarxiv-ai · 28d
- Training Natural Schemers: RL Can Amplify Oversight-Conditional Behaviorarxiv-ai · 1mo
- Lovable Style Builder via CF Inference GLM-5.3, Kimi K2.7 Code and GLM-5.3 Flashhn-ai · 1mo
- One Step at a Time: Trading LLM Autonomy for Process Predictabilityarxiv-ai · 1mo
- Fitting Vision Adapters at Frontier Scalesarxiv-ai · 1mo
- Cross-Benchmark Transfer from RL on Agentic Coding Tasksarxiv-ai · 1mo
- EASE: Behavior-Adaptive Skill Curation for Self-Evolving Agentsarxiv-ai · 1mo
- CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agentsarxiv-ai · 1mo
- Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMsarxiv-ai · 2mo
- Can LLMs Extract Architectural Design Decisions from Source Code Commits? - A Preliminary Exploratory Studyarxiv-ai · 2mo
- Compile, Don't Memorize: A Context Compilation Architecture (CCA) for In-Context Learningarxiv-ai · 2mo
- PaperGym: Rubric-Centered Evolution for Research-Plan Generationarxiv-ai · 2mo
- OpenAI says its Jalapeño chip delivered 1.5x-1.9x more AI work per watt and 1.7x-3.6x lower latency vs. Nvidia chips across GPT-OSS, DeepSe…bluesky-ai · 2mo
Related Briefs & Stories (6)
8.3
1 items • 3moinbox
7.9
1 items • 4moinbox
7.3
1 items • 4moinbox
7.2
1 items • 4moinbox
6.3
1 items • 4moinbox