Milestones
The story so far.
Qwen3.8 27B goes live
Upgraded from Qwen3.6 27B — now with reasoning effort per request, from zero thinking to full deliberation.
Kokoro Text-to-Speech drops
Text, vision, embeddings, speech-to-text — and now text-to-speech with natural voice replies. 68 voices, 10 languages, ~74× realtime.
DeepSeek-V4-Flash upgraded to 0731
Massive intelligence boost — the v4 series keeps getting smarter.
Whisper Speech-to-Text drops
Text, vision, embeddings and now speech-to-text — the lineup grows.
Vision augmentation added
Qwen3.8 now powers image analysis for DeepSeek-V4-Flash — bringing vision capabilities to both models.
DeepSeek-V4-Flash goes live
Replaced Gemma-4 31B with 284B MoE (13B active) — dramatically more capable.
Gemma-4 & Qwen3-Embedding added
Infrastructure expanded with additional models — broader capability, same flat price.
Customers from 25+ countries
Swiss-hosted, globally adopted. From zero to worldwide in one month.
Business accounts launched
Multi-seat management and custom pricing.
Qwen3.6 27B goes live
Upgraded from Qwen3.5 122B A10B — smarter, sharper, same price.
AI Router went live
Day one. Unlimited Qwen3.5, Swiss-hosted, flat pricing.
Waiting list online
Collecting early interest — the very beginning.
Join the journey
Be part of the story. Build the next milestone yourself.
Subscribe Now arrow_forward