Search results for Claude Sonnet
AI & Enterprise
Testing ChatGPT-6 Astra for tedious tasks from train tickets to account deletion
A reporter tested OpenAI’s GPT-6 Astra on delayed web chores ranging from an OTT compensation claim and account deletion to bargain shopping and Chuseok train tickets. It could navigate sites, summarise conditions and complete multi-step forms, but logins, phone verification, CAPTCHAs and anti-macro blocks limited hands-free use. It performed well on checking National Assembly schedules and creating a GIF thumbnail, and it helped upload images and captions in a newsroom editing system.
AI & Enterprise
AI search startup Glean says tokens 70 percent cheaper than Claude CoWork - how?
AI search startup Glean is stepping up its offensive against Anthropic, targeting token costs and data security issues. It released an internal evaluation saying its AI assistant was far cheaper than Anthropic’s Claude CoWork app when performing more than 180 tasks. Glean said it used 70 percent fewer tokens and its average charge per task was 81 percent lower, citing routing software that assigns tasks to different models.
AI & Enterprise
Google, OpenAI race to ultrafast AI with Gemini 3.7 Flash and GPT-5.6 Sol Ultrafast
Google and OpenAI unveiled ultrafast AI services targeting coding and AI agent work, highlighting a shift in competition toward response speed. Google launched the low-cost Gemini 3.7 Flash in more than 160 countries, with support for up to 1 million input tokens and 64,000 output tokens and multimodal processing. OpenAI introduced GPT-5.6 Sol Ultrafast, a faster processing tier for GPT-5.6 Sol using Cerebras chips, initially limited to some API customers.
-
AI & Enterprise
Cloudflare automates bug bounty processing with AI for $58 a month
-
AI & Enterprise
Claude Code sends 33,000 tokens before processing prompts, rising to 75,000 in real use
-
AI & Enterprise
Neocloud market shifts as SoftBank and Meta enter
-
AI & Enterprise
Anthropic unveils Claude Sonnet 5, lowers price for Opus 4.8-level performance
-
AI & Enterprise
Xiaomi research team unveils \'HarnesX\' auto-evolution framework for AI agent harnesses
-
AI & Enterprise
Chinese AI models overtake U.S. in token usage, DeepSeek V4 ranks No. 1
-
AI & Enterprise
Upstage CEO Kim Sung-hoon says Fable5 foreign access ban highlights need for Korean sovereign AI
-
AI & Enterprise
GitHub Copilot fully implements token-based billing as users complain monthly limits shrink fast
-
AI & Enterprise
Meta developing \'Hatch\' personal AI agent, also targeting shopping agent
-
AI & Enterprise
Meta\'s open-source Llama effectively abandoned; what are developers\' alternatives?
-
AI & Enterprise
DeepSeek launches new V4 model preview versions
-
AI & Enterprise
Why Google Gemma 4 is in the spotlight as an open-source AI game changer
-
AI & Enterprise
LG unveils EXAONE 4.5 multimodal AI model focused on stronger reasoning
-
AI & Enterprise
Anthropic says Claude Sonnet 4.5 test drew blackmail plan after replacement email
-
AI & Enterprise
Chinese AI overtakes U.S. in token consumption, expands global share on price competitiveness