4/ Search costs in the same benchmark: $9.07 per 1,000 tasks for Octen versus $65.57 for Exa auto. Roughly one-seventh. That's the search bill. Including model costs, the totals are $58.22 and $127.15 respectively.
ECONOMY
-
Octen Leads a Search Speed Benchmark at 0.21 Seconds
By
–
3/ Octen Search (highlights) scores 77, tied for third on AA's Search Index. It also leads on speed: 0.21 seconds per search query and 15.6 seconds per task, including AA's estimated model time. Board data: September 8.
-

Octen Search Benchmarks Agent Query Speed, Quality, and Cost
By
–
1/ Octen gives AI agents live web search at 0.21 seconds per query in Artificial Analysis' benchmark. It is the only Search API in the top 3 for quality, speed and search cost. Search fees: roughly 1/7 of Exa auto's. Sept 8 data. @OctenAI @KZouAPT
-
Google’s Project Suncatcher Plans to Test TPUs in Orbit
By
–
-
AI Agents Could Change the Economics of Advertising and Messaging
By
–
I think the ad based internet is wrong from an incentives perspective, and it may not be compatible with AI augmented user agents. The correct way seems to be paying the cost of recipient side filtering for sending messages, and recouping that cost from willing recipients.
-

Blackhole Galaxy Reports Faster LTX Video Inference Without Compression
By
–
Blackhole Galaxy® runs LTX-2.3 Fast faster than anyone else — full model, zero compression, no quality tradeoffs, at over 2x the speed of other inference providers. Check it out on @ArtificialAnlys
: https://
artificialanalysis.ai/video/provider
s/ltx-2-3-fast
… -
Inference Speed Matters Once Token Spending Produces Positive Returns
By
–
I would expect this too 👀 As soon as you find a way to have a positive ROI on token spend, the speed would make a huge difference for the business case growth/market expansion.
-

An AI Inference Discussion Starts With Hardware and Workload Shape
By
–



If you cannot read a quoted tweet as context then I don’t know what to say. I quote-tweeted all the context, so don’t be disrespectful. I am teaching people about kernels and optimizations everyday, so in the context of my content it’s also signaling.
-

Anthropic’s Co-Founders Could Retain Majority Voting Control After IPO
By
–
Anthropic reportedly plans to give its seven co-founders majority voting control ahead of its IPO, which gives them much more power. The Information reports that a proposed Palantir-style share structure would give them 50.1% of voting power on most corporate matters, as long as
-

Mac Studio and DGX Spark Trade Leads in Inference Benchmarks
By
–
Most notable thing M5 Ultra with 1.2 TB/s bandwidth is only 6.4% faster in token generation than the DGX Spark (273 GB/s) M5 Ultra has ~5x the Bandwidth of the DGX Spark