IndustryBaseten·9d ago·3 min readSummaryListen 0:47
Introducing Baseten Hosted ToolsBaseten introduced hosted tools for deploying and managing AI models.
News Hub
What actually shipped in agent engineering, pulled from the labs, arXiv and Hacker News.
See who we follow52 articles
ToolingBaseten·14d ago·3 min readSummaryListen 0:49
DeepSeek-V4.1-Flash: more efficient prefill for coding agentsDeepSeek-V4.1-Flash improves prefill efficiency for coding agents.
Also covered byFireworks AI
ToolingBaseten·15d ago·3 min readSummaryListen 0:36
Blaxel is joining Baseten to build the future of agentic infrastructureBlaxel is joining Baseten to build agentic infrastructure.
ResearchBaseten·17d ago·3 min readSummaryListen 0:48
Baseten leads Coval’s voice AI benchmarkBaseten leads a benchmark for voice AI systems.
ToolingBaseten·23d ago·3 min readSummaryListen 0:48
New MCP and skill for coding agents to use BasetenBaseten released a new MCP and skill enabling coding agents to use Baseten's platform.
The efficient frontier of LLM inference
Article discusses optimization techniques for reducing computational costs during language model inference.
ToolingBaseten·28d ago·3 min readSummary
Agentic kernels in productionAgentic kernels are now deployed in production systems.
ToolingBaseten·28d ago·3 min readSummary
GLM 5.3: Scaling with post-training, intuitively explainedGLM 5.3 scales model performance through post-training techniques.
Open sourceBaseten·29d ago·3 min readSummary
Best open-source models for post-trainingArticle discusses best open-source models available for post-training.
ToolingBaseten·29d ago·3 min readSummary
fastest video inference, everArticle reports fastest video inference performance achieved.
ToolingBaseten·31d ago·3 min readSummaryListen 0:55
The two AI gateway patterns in production inferenceBaseten describes two AI gateway patterns used in production inference systems.
ToolingBaseten·32d ago·3 min readSummaryListen 0:40
How to run any open model inside DeepSeek HarnessDeepSeek Harness enables running any open model inside its framework.
ToolingBaseten·32d ago·3 min readSummaryListen 0:42
How leading platforms ensure observability for LLM inferenceLeading platforms ensure observability for LLM inference.
ToolingBaseten·43d ago·3 min readSummaryListen 0:43
Inference engineering for DeepSeek V4 Pro 0813Baseten provides inference engineering guidance for DeepSeek V4 Pro 0813.
Open sourceBaseten·43d ago·3 min readSummaryListen 0:54
Baseten delivers open-source inference for You.comBaseten provides open-source inference capabilities for You.com.
ToolingBaseten·44d ago·3 min readSummaryListen 0:42
Qwen3.8-Max: Alibaba's new frontier reasoning modelAlibaba released Qwen3.8-Max, a new frontier reasoning model.
IndustryBaseten·52d ago·3 min readSummaryListen 0:45
Introducing NVIDIA Nemotron 3.5 ASR StreamingNVIDIA introduced Nemotron 3.5 ASR Streaming for automatic speech recognition.
ToolingBaseten·53d ago·3 min readSummaryListen 0:47
Laguna S 2.1 goes Greek: a repository-scale game transformationLaguna S 2.1 enables repository-scale game transformation.
ToolingBaseten·56d ago·3 min readSummaryListen 0:42
Fine-tuning Qwen3-TTS for high-quality voice cloningBaseten demonstrates fine-tuning Qwen3-TTS for high-quality voice cloning.
ToolingBaseten·57d ago·3 min readSummaryListen 0:48
22,580: GPT-2 to Kimi K3, explainedBaseten explains model progression from GPT-2 to Kimi K3.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.