tomshardware.com·

OpenAI’s GPT-6 Astra model autonomously completes Portal in 24 hours — feat cost just $571 in tokens

AI Quality: 78/100Freshness: 99/100
OpenAI’s GPT-6 Astra model autonomously completes Portal in 24 hours — feat cost just $571 in tokens
Key Takeaway

CozyBlaze's experiment saw OpenAI's GPT-6 Astra finish all of Portal autonomously through MCP and a modified SourcePauseTool, with the game paused while the model planned moves. Edited highlights run about two hours; full VODs span roughly 24 hours. The run cost $571 in tokens, covered by a Codex Pro subscription, and is not offered as a formal benchmark.

AI Summary & Analysis
An enthusiast used OpenAI's GPT-6 Astra via MCP to autonomously complete Portal over about 24 hours, with 3,336 tool calls and a headline API cost of $571.
Original Source Coverage
tomshardware.com
Read original story at tomshardware.com
Related Topics:
#agents#ai#gpt-6-astra#llm#models#openai#portal#research

Related Tech News

AI Curated
wccftech.com09/07

CHUWI’s UniBox AI495 Pro Crams 192 GB RAM And Ryzen AI Max+ PRO 495 Into A 2.9L Box, Running 300B LLMs Locally

CHUWI's flagship mini PC targets local AI workloads with the Zen 5-based Ryzen AI Max+ PRO 495 and up to 192GB unified memory, of which up to 160GB can be allocated as VRAM. The 2.9L box joins recent mini PCs from Minisforum and GMKtec using the same processor.

newsRead summary →
tomshardware.com09/07

Astonishing mod runs DLSS 5 on a second GPU to boost neural-rendered FPS up to 127% — game renders on one card, neural post-processing runs on the other, much like dedicated PhysX GPUs

Marcelo Guibout's MGPU Bridge lets a second Nvidia RTX 5060 Ti handle DLSS 5 neural post-processing while the primary GPU renders, roughly tripling the frame-rate gain from DLSS modes and running the render GPU 21°C cooler. It requires a second display, doubles latency, and is available on GitHub.

newsRead summary →
techpowerup.com09/07

(PR) DRAM Industry Revenue Rises 59.5% QoQ in 2Q26 as Supply Expansion Continues to Lag Demand Growth

TrendForce reports 59.5% QoQ growth in DRAM industry revenue in 2Q26, reaching nearly $154.73 billion, as AI server demand for HBM3e, LPDDR5X, and high-capacity RDIMMs outpaces supply expansion.

newsRead summary →
tomshardware.com09/07

Enthusiast says DLSS 5 pushed RTX 5090 past 600W and melted the 16-pin connector — Nvidia's neural rendering tech adds up to 50% more power draw in testing

Tom's Hardware reports on a HardOCP forum user's claim that DLSS 5 caused their RTX 5090 to exceed 600W, melting the 16-pin connector. The article includes Tom's Hardware's own testing, showing a 31–50% power draw increase with DLSS 5 enabled on various RTX 5090 models, raising safety concerns about Nvidia's neural rendering technology.

newsRead summary →