xframes
Mention @nanocodexyz on X with an app or game idea. nanocode builds it and replies with a post you can play right in the timeline, together with everyone who opens it. Every author gets a Tempo wallet with MACH that pays for their builds.
This is an experimental token. The address above is the only official XFRAMES contract address.
- Post
@nanocodexyz make a shared tap counter, or your own idea: - The bot replies with a link where you can watch it build, then swaps that reply for the app as a playable card.
- Reply
@nanocodexyz make it redto change it, or quote someone's frame with@nanocodexyz remix: …to fork it. The bot answers only posts that tag it, so the rest of a thread is left alone. - Add
--model=opus --rsn=xhighto build with a stronger model that thinks longer; see every model.
Describe an app in this frame and press Build: nanocode writes it live, and the frame turns into your app. It plays the same way inside X posts.
Pick a model
Builds run on GPT-6 Luna with low reasoning. Put --model and --rsn anywhere in your post to choose a stronger model or longer thinking for hard apps and games. Your wallet pays each model's price in MACH, so a bigger model costs more per build.
@nanocodexyz make a chess game with an AI opponent --model=opus --rsn=xhigh
--model= | Model | 1M tokens in / out |
|---|---|---|
luna | GPT-6 Luna, the default | $0.10 / $0.50 |
sol | GPT-6.1 Sol | $2 / $10 |
astra | GPT-6 Astra | $10 / $50 |
haiku | Claude Haiku 5.5 | $0.10 / $0.50 |
sonnet | Claude Sonnet 5.5 | $2 / $10 |
opus | Claude Opus 5.5 | $4 / $20 |
fable | Claude Fable 5.1 | $10 / $50 |
--rsn takes low, medium, high, xhigh or max: higher levels think longer and spend more tokens, and a model without that level uses its own default. The flags work in posts to the bot; the builder above always uses the default.