Skip to main content
usedby

50%

“the 50% cost reduction compared to traditional cloud providers”

As published on together.ai. Checked by usedby on Oct 9, 2026.

What happened

SCB 10X trains its Thai-language Typhoon models on Together AI's reserved GPU clusters and serves them through Together AI's serverless inference endpoints as Typhoon API Pro.

Summary written by usedby from the source page, in English. The figures are those of Together AI and SCB10X, not ours.

Together AI provided SCB 10X with customizable bare metal clusters featuring dedicated NVIDIA GPU hardware. The Together Kernel Collection delivered low-level optimizations that supercharged training speeds, enabling the team to run multiple experiments concurrently and efficiently.

From the page. together.ai

What the story claims, and what we checked

What we compared with the page.

  • The figure: 50%CheckedPrinted word for word on the page, near the name of SCB10X.
  • The passage quoted aboveCheckedCopied word for word from the page, near the name of SCB10X.
  • The numbers in our summaryCheckedEach one is printed on the page.
  • SCB10X uses Together AICheckedConfirmed line.
  • The result itselfNot checkedWe quote it; we did not measure it.

Same company, same tool or same industry.

95%“Time to First Token (TTFT): Improved by up to 95%”Arcee AI uses Together AI. Another customer of Together AI6דDecagon achieved nearly 6× cost reduction per turn compared to closed models like GPT-5 mini.”Decagon uses Together AI. Another customer of Together AIunder two weeks“Together’s team shipped FP8, FP4, and INT4 quantized variants of the Cogito models in under two weeks”Deep Cogito uses Together AI. Another customer of Together AI