Skip to content
Drix10 Blog

Tether market value share vs conversation share

, 4 items in AI for Content Creation and Marketing, 3 min read

In this digest (4 items)

Tether represents 6.6 percent of crypto’s market value. It generates only 0.54 percent of crypto‑related conversation. USDC holds 2.6 percent of both market value and conversation.

Key points

  • Market share: Tether 6.6 percent of crypto’s market value.

  • Conversation share: Tether 0.54 percent of crypto talk.

  • USDC share: USDC 2.6 percent of market value and conversation.

Sources

Delta earnings release – $1.72 per share before open

Delta ($DAL) posted adjusted earnings of $1.72 a share before the market opened. Shares were down about 3.3% in pre‑market trading. Sentiment moved from 90 to 78. Revenue grew.

Key points

  • Adjusted earnings: $1.72 a share before the open

  • Shares down: about 3.3% pre‑market

  • Sentiment change: 90 to 78

  • Revenue grew​

Sources

OpenAI Decision Model Router launches with Luna model

OpenAI’s new Decision Model is powered by Luna and is accessible via the Swarms Decision Model Router. Developers can switch to Jev by typesafeai or Clef by Cloudflare with a single line of code. The Decisions API claims up to 10× faster decisions than Luna through the Responses API and is in public beta.

Key points

  • Decision Model Router provides access to Luna model

  • Switch to Jev (typesafeai) or Clef (Cloudflare) with one line of code

  • Decisions API offers up to 10× faster decisions than Luna via Responses API

  • Decisions API is available to all developers in public beta

Sources

PressBot benchmark for WordPress Admin Agent

PressBot released a benchmark of 11 judgment tasks run three times per model, grading on actual site changes. The benchmark found bugs in PressBot’s own agent and showed the cheapest Claude model scored perfectly. Claude Haiku 5.5 is now the default Admin Agent model at $0.0063 per task.

Key points

  • Benchmark: 11 judgment tasks, 3 runs per model, graded on site changes

  • Results: All three Claude models passed every task; Claude Haiku 5.5 recommended default at $0.0063 per task

Sources