Together rewrites its inference platform after serving 400 trillion tokens a month
graceisford · x · 2026-07-24
Together says it has rewritten its inference platform from the ground up after learning how AI-native teams and enterprises use open and custom models in mission-critical production.
What changed
- The new platform is now live.
- It is designed for running open models in production with full control.
- Teams can test every change on live traffic before users see it.
- The company says it serves 400 trillion tokens per month, which shaped the redesign.
The focus here is infrastructure: safer production workflows, controlled deployments, and operational lessons from large-scale inference traffic.
Related event: Together.ai Launches New Open Model Inference Platform(4 posts)→
More from Infra
- Engram's random reads don't suit SSDs; CPU-memory over NVLink could serve all 72 GPUs — bookwormengr · 2026-09-11
- 80% of the DIY LLM inference hype posters have already quit — it's brutally hard systems work — abhijithneil · 2026-09-11
- Hugging Face's Ultra Scale Playbook: a free book on training LLMs on GPU clusters — mdancho84 · 2026-09-11
- Is inference latency becoming the biggest bottleneck for production AI agents? — Euphoric_Sea632 · 2026-09-11
- LLM Serving Metrics Thread: Why TPOT and Uptime Make or Break User Experience — abhijithneil · 2026-09-11
- PlanetScale launches sharded Postgres: 768 servers acting as one, 1PB scale — dhruv2038 · 2026-09-11