OneTriangleAI launches ultra-low latency DeepSeek V4 hosting
ycombinator · x · 2026-08-25
YC S23 startup OneTriangleAI has launched DeepSeek V4 Flash hosting at aggressive rates ($0.15/M input, $0.35/M output). Founded by MIT grads and Olympiad medalists from DeepMind and Jane Street, the service sets new standards for cost and latency via KV cache optimizations.
More from Infra
- AI Data Centers Face Local Resistance: Jobs Needed as Energy Costs Rise — dbasch · 2026-08-25
- SpaceX Space Data Center: Leveraging Starlink for Power and Cooling — XFreeze · 2026-08-25
- How to Subscribe for Access to Tons of Open Weights Models? — chrisalbon · 2026-08-25
- Nvidia spending $6B to build US alternative to Chinese AI — abhishekcode42 · 2026-08-25
- NVIDIA's Vera CPU charts criticized for misleading comparisons against older AMD chips — zephyr_z9 · 2026-08-25
- How good is Fujitsu Monaka CPU? — firstadopter · 2026-08-25