baseten

Bloomberg 报道 AI 推理基础设施公司 Baseten 将融资约 20 亿美元，估值最高达 300 亿美元，较六月估值翻倍以上；其年化收入据称在 2026 年第一季度从 2 亿美元跃升至 6 亿美元，反映出推理需求的快速增长。

The author recommends Baseten's "Inference Engineering" book, suggesting readers with some background start with Chapters 5, 6, and 2 to systematically learn about LLM inference techniques.

OpenAI enterprise customers can now apply their existing usage commitments toward Moonshot AI's Kimi K3 model through Baseten's deployment platform, broadening cross-vendor model access for business users.

Goodfire launched 'inside-out' AI safety monitors that read a model's internal activations rather than rereading its outputs, catching rogue agents at far lower cost and latency. Available to Baseten customers, the probes detected 94% of malicious hacking sessions at roughly $51 per 1,500 sessions versus $10,000 for a top-tier AI monitor.

OpenAI launched a B2B enterprise marketplace at DevDay with 32 partners, letting customers apply existing OpenAI commitments to third-party products including open models served by Baseten — a strategic move to become the hyperscaler layer for AI apps as open-weight models grow in volume.

Baseten announces a partnership with OpenAI, making Baseten-powered open models available to OpenAI customers so organizations can build multi-model agentic coding workflows, routing tasks to the best-fit model across intelligence, cost, latency, and capacity.

Baseten CEO Tuhin Srivastava reveals that token volume on Baseten has grown 40x year-over-year, with revenue increasing about 10x in the last 12 months, highlighting the explosion in AI inference.

Strix, an autonomous hacking agent, discovered and reported a critical security vulnerability in Baseten's GitHub, gaining admin access through an exposed token that was quickly fixed by Baseten's team.

The article recommends a YouTube episode featuring Charlie O'Neill from Baseten, discussing how Kimi and GLM models are superior to Opus 5 and addressing skepticism about AI progress and AGI.

Niels Rogge adds web search to Papers with Code's upcoming chat interface, powered by Hugging Face Inference Providers, Baseten's DeepSeek-V4 Flash, Exa's MCP server, and Modal. He invites HF usernames for feedback.

Hugging Face announces Baseten as a supported Inference Provider, enabling serverless access to popular open-weight LLMs like Kimi K3, DeepSeek V4 Flash, and GLM-5.2 directly from model pages and SDKs.

Baseten details how it built the fastest API for GLM-5.2, achieving over double the launch-day performance and introducing a latency-optimized Fast version for coding and agents, with further improvements planned.

Baseten announces the world's fastest API for the GLM-5.2 open model, achieving over 280 tokens per second via NVFP4 quantization, disaggregated inference, and other optimizations.

Baseten processes over 1 billion inference calls per day and has raised $1.5B to scale its infrastructure, highlighting inference as a massive market.

Apoorv Agrawal from Altimeter Capital explains why they are doubling down on their investment in Baseten, arguing that inference will become the largest market and that post-trained open source models offer the best combination of capability, cost, and control.

Baseten announces a $1.5B Series F funding round led by multiple investors, citing 20x revenue growth and 40x inference volume growth as evidence of the market's shift toward inference as the key AI layer.

Baseten, a $13 billion AI startup, provides software and computing capacity to companies using lower-cost AI models as alternatives to OpenAI and Anthropic.

A tweet from Philip Kiely highlights cost savings by switching from closed-source AI models to open-source alternatives, using Baseten's ROI calculator tool.

Cursor AI launches a new interview series with developers, starting with a conversation with the Baseten team about their use of coding agents, current workflows, and future predictions.