Please confirm you are human

This browser or connection looks automated. Press and continuously hold the control for 3 seconds to enable Google-hosted web results and, when separately allowed, AI-assisted answers.

A successful check enables 100 search requests. Interactive access does not authorize scraping, systematic collection, or reuse of search output.

Hold with a pointer, or hold Space or Enter.

News

The New Stack
thenewstack.io > openai-agents-api-compute

OpenAI’s researchers burned $7,000 a day on AI agents — now it’s opening the floodgates

2+ day, 12+ hour ago   (155+ words) OpenAI's Agents API sells the Codex harness as a service, making long-running agents easy to build — just as Astra demand forces a Pro plan pause....

The New Stack
thenewstack.io > gpt-live-1-voice-api

OpenAI split a voice model's brain. Then one team deleted 23,000 lines of code.

3+ day, 14+ hour ago   (206+ words) OpenAI's GPT-Live-1 API splits voice agents into a fast conversational model and a separate reasoning backend, cutting one customer's codebase by 80%....

The New Stack
thenewstack.io > claude-build-agents-benchmark

Claude did best on a new benchmark for agents that build agents. It still passed fewer than a quarter of the tests.

4+ day, 14+ hour ago   (492+ words) Sierra has open-sourced Hyper-𝜏-bench, a follow-up to its 2024 τ-bench that tests how well AI agents can build other agents....

The New Stack
thenewstack.io > k2-horizon-fully-open

K2 Horizon just shipped as six new fully open models — developers aren't fully convinced

4+ day, 22+ hour ago   (632+ words) Based in the Emirati capital, Abu Dhabi, the Institute of Foundation Models (IFM) introduced K2 Horizon last week. This group of six AI foundation models, ranging from 0.9 billion to 375 billion parameters, is claimed to be the “largest fully open-source fleet of…...

The New Stack
thenewstack.io > deepseek-dsec-agent-hiring

DeepSeek is hiring 150 engineers, and none of them will touch a model

5+ day, 15+ hour ago   (172+ words) DeepSeek announced 150 backend engineering roles to scale DSec, the sandbox system running hundreds of thousands of concurrent agent environments on a single cluster....

The New Stack
thenewstack.io > astra-reasoning-effort-cost

OpenAI's new model costs 2.5x more per token — and developers are saving money anyway

6+ day, 18+ hour ago   (300+ words) OpenAI says GPT-6 Astra at low reasoning beats Sol at high — and can cost less per task despite 2.5x token prices. Here's what the benchmarks show...

The New Stack
thenewstack.io > gpt6-astra-developer-access-delayed

"Sorry for the messy rollout": OpenAI launched GPT-6 Astra, but developers are locked out

1+ week, 2+ day ago   (29+ words) OpenAI launched GPT-6 Astra but most developers still can't access the API. Here's what went wrong with the rollout and when broad access may arrive....

The New Stack
thenewstack.io > coder-cursor-agent-relay

"1% of my engineers are responsible for 40% of token spend": Why Coder and SpaceXAI want to give developers nice things

1+ week, 2+ day ago   (145+ words) Coder now lets regulated enterprises run Cursor's cloud agents on their own infrastructure, without losing security control....

The New Stack
thenewstack.io > astra-arc-agi-benchmark

GPT-6 Astra aced the hardest AI benchmark. The asterisk matters more than the score.

1+ week, 3+ day ago   (217+ words) OpenAI says GPT-6 Astra scored 98.6% on ARC-AGI-3, up from 7.8% six months ago. But undisclosed test settings and opaque reasoning complicate the AGI question....

The New Stack
thenewstack.io > openai-gpt6-astra-benchmarks

OpenAI launches GPT-6 Astra and says welcome to the "AGI era"

1+ week, 3+ day ago   (1229+ words) OpenAI’s new flagship posts big gains on specialized tasks, but it comes at a premium and does not clearly lead the coding pack....