
LFM2.5-24B-A2B on the Qtum Router: an agent model at $0.03 per million output tokens
Liquid AI's new agent model is live on the Qtum Router at 10 credits per million input tokens and 30 per million output — a tenth to a…
Read →Writing from the Qtum.AI team on AI model routing, LLM gateways, AI video generation, API portability and what inference actually costs.

Liquid AI's new agent model is live on the Qtum Router at 10 credits per million input tokens and 30 per million output — a tenth to a…
Read →
Flagship video models charge about nineteen times what a draft model does per second. Lock the look on stills, prove the camera move at…
Read →
DeepSeek V4 Flash, MiniMax M2.7 and GLM 5.3 Flash now run on decentralized compute — a flat 6 credits per million tokens, input and output…
Read →
Two environment variables put Claude Code on the Qtum AI Router: one prepaid balance, every Claude model, and a per-request record of what…
Read →
AI video pricing is usually a pack of credits and a shrug. Here it is one multiplication you can do before you spend anything - and the…
Read →
Point the OpenAI or Anthropic SDK you already use at one endpoint and reach 91 models - text, image, audio and video - on a single prepaid…
Read →
Shot structure, the dozen lens and lighting terms that do most of the work, and four prompts rendered on two models each so you can see…
Read →
One plugin, one auth profile, one verification call — and every model in the Qtum catalogue becomes a native OpenClaw model id you can…
Read →Most of what we write about here is what the Qtum AI Router and Qtum Video were built to solve.