Access ChatGPT and all other text and image models. Anonymous, private, pay-per-prompt using Nano, other crypto, or credit cards!
Joined January 2024
- Tweets2.6K
- Following2
- Followers3.9K
- Likes4.3K
We quietly hid a treasure hunt inside nano-gpt.com 💎
7 clues. 7 words. 1 secret vault.
Someone cracked it and won $50 in credits 🏆
The prize is gone, but the diamond is still there.
AI agents mostly got stuck. Can you beat them?
Start where developers look first.
Related: we're seeing a massive uptick in uncensored and abliterated model usage, along with a big increase in DMs and emails about how to red-team your codebase.
Red-teaming from inside the codebase is, luckily, far easier than from the outside, so get it done!
For all those asking how best to red-team your own codebase we now have an easy setup.
It uses the amazing @t3dotcodes with the @abliteration_ai models (GLM 5.3 & Deepseek V4.1 Flash).
nano-gpt.com/api?tool=t3#cod…
For all those asking how best to red-team your own codebase we now have an easy setup.
It uses the amazing @t3dotcodes with the @abliteration_ai models (GLM 5.3 & Deepseek V4.1 Flash).
nano-gpt.com/api?tool=t3#cod…
Chat prompt presets can now save a final reminder after the conversation history.
Reinforce format, tone or language, reuse it in new chats, and keep it through preset import/export and cross-device sync.
nano-gpt.com/chat
NanoGPT's Usage overview now shows your input/output token split.
See percentages alongside token totals, and hover over the charts for the split in each time period. Easier to spot input-heavy workloads.
nano-gpt.com/usage
P-Video 2 Pro now has a Talking Avatar mode on NanoGPT.
Upload a portrait and 1-180 seconds of speech to make it talk, at 480p or 768p. Select Talking Avatar in the generation controls.
nano-gpt.com/studio/video?mo…
Kling 4.0 Preview is now on NanoGPT.
Create 3-30 second videos at 720p or 1080p. Start from text or images, guide motion with references and keyframes, edit footage or extend a clip.
nano-gpt.com/media?mode=vide…
NanoGPT can now package generated code into a downloadable ZIP.
Enable ZIP archive in Tools, then ask for a multi-file project. Files keep their folder structure, so you don't have to copy each one separately.
nano-gpt.com/chat
Kimi K3.1 Pro and Pro Highspeed are now on NanoGPT.
1M-token context, text/image/video understanding, and tool calling for coding, document analysis and agents. Choose low, high or max reasoning.
nano-gpt.com/conversation?mo…
PixVerse V6 Max is now on NanoGPT.
Turn text or a starting image into 1-15 second videos at up to 1080p. Add native audio, enable multi-clip generation, or set a seed.
nano-gpt.com/media?mode=vide…
HeyGen Voice and Voice Turbo are now on NanoGPT.
Choose from 24 voices or clone a speaker from a reference recording. 39 languages, adjustable expressiveness and 44.1 kHz WAV output, with up to 5,000 characters per request.
nano-gpt.com/studio/audio?au…
Also live on NanoGPT!
Now live: abliterated-model-large-flash (our version of DeepSeek-V4.1-Flash-1017). It is a cheaper alternative to our GLM-5.3-based model for product owners and defenders who need access to the most cyber-capable models.
- 1M context
- Vision
- Zero input/output retention or training
We believe paying customers should be in control, and over the past few months, leading experts and major brands have adopted our models.
abliteration.ai
SubQ Snap 1.1 is now on NanoGPT's Decisions API.
Classify, score and route text, JSON or inline images with typed answers and probabilities. 384K context, included in subscriptions at the regular 1x rate.
nano-gpt.com/models/decision…
Abliterated Large Flash is live on NanoGPT.
Uncensored reasoning based on DeepSeek V4.1-Flash-1017, with image input, tool calling, structured JSON output and a 1M-token context.
nano-gpt.com/conversation?mo…
Ternary Bonsai2 9B is now on NanoGPT.
Prism ML's open-weight model supports reasoning and tool calling with a 262K-token context. Thinking is on by default, and can be turned off.
nano-gpt.com/conversation?mo…
FLUX.3 Max is now on NanoGPT.
Create 5-20 second videos with native audio at 720p or 1080p. Start from text, a start/end image pair, or up to 10 ordered keyframes.
Sound on for this provider sample.
nano-gpt.com/media?mode=vide…
Ideogram Vector is now on NanoGPT.
Generate downloadable SVG graphics from text, with up to five reference images. Or convert a PNG, JPEG or WebP into SVG.
Three quality levels. Up to eight results per generation.
nano-gpt.com/media?mode=imag…
Turn a video into a narrated ad, tutorial or explainer in NanoGPT Studio.
Paste a script, import timed lines, or let AI draft one. Pick a voice, adjust the timing, and keep or replace the original sound.
For videos up to 3 minutes.
nano-gpt.com/studio/audio
MiniMax M3.1 Flash is now on NanoGPT, included in your subscription.
Built for coding and tool use, with text, image and video input, a 1M-token context, and always-on reasoning.
nano-gpt.com/conversation?mo…
These models (luckily!) work far better when they can see the codebase they're working on, so there is some defender's advantage here.
We're getting quite some questions about what model to use to red team your own codebase.
Claude and Astra are fantastic but run into guardrails often, Abliteration AI Large V2 (GLM 5.3 with guardrails removed) is seeing a surge in usage.
nano-gpt.com/models/text/abl…
