Author of Crawl4AI. AI & software builder, systems thinker, founder of Kidocode. I build, analyze, and advise on hard technical problems.
Singapore
Joined September 2009
- Tweets2.8K
- Following376
- Followers3.1K
- Likes3.7K
Another good use I found for @typesafeai’s Jev: automatically switching Claude Code models based on what I’m working on.
I built JevShift for this. It looks at the task and recent context, then picks a model. Here’s a quick demo 👇
Code and installation: github.com/unclecode/jevshif…
Next I want to do the same for thinking effort, so Jev can pick both the model and how much thinking the task needs.
i fingerprinted this "union alpha" which turned out it is (unbiased/pareto) for 15 times in 100 minutes:
02:05-02:34 UTC: a fingerprint matching hermes-4-405b in my library, 4 of 4 probes.
02:37: every probe value changes at once. new model.
02:40-03:46: a signature matching nothing i have measured. unknown model.
and in 10 of the 15 runs, the same text sent twice, seconds apart, was billed with two different token counts.
people spent the stealth week arguing whether it was openai, llama, qwen or glm. everyone was right. they were fingerprinting different models behind one name.
Not only for mystery models. Your daily OpenAI or Gemini API deserves the same check. modelprint.ai
Union Alpha is revealed on OpenRouter as @TheUnbiasedCo Pareto!
Impressive detective work and demand from the community cut this stealth period short, but you can continue using it now: openrouter.ai/unbiased/paret…
This is how I matched: modelprint fingerprints 23 known models every day, so when a mystery signature shows up i just compare. modelprint.ai/history
btw i also measured the models the viral post says are inside pareto. none of them match what it served me. rumors need measurement too :)
Unclecode (Hossein) retweeted
New in Claude Code: claude plugin eval
See what value your plugin is adding, or if it needs more work.
You can create test cases, run your plugin or skill against those test cases, score those runs, then run each case again without the plugin to see the differences.
I really wonder how will be the saas model? Here @treg_ai let us access to 2800+ saas and pay as u go!
Grok bot just killed 2800+ SaaS...
Instead of paying $600 subscriptions, build a loop with Treg
Try it at treg.to/grokbot
Cost cents per call, and tailored to just your needs
No subscription, 0% markup
Fully open sourced, git repo below 👇
Fabel 5.1 is where finally I feel claude code talk concise and relevant! It gets the point much faster, do not generate long essay result for a simple question and so far so good!
I think paying for all LLM should become pay as you go! Pay only when u use and not monthly subscription! Unless u r a heavy users. Is that something u look? Check this treg oss library and hosted version
Fable 5.1 + Treg = Run GTM fully in a terminal
No more $200 subscriptions, just $0.0004/call
Fable 5.1 excels at complex, long-running tasks + @treg_ai connects it to 2800+ data & tools:
Leads enrich, SEO, social, ads
Fully open sourced
No subscription, 0% markup
Try it at treg.to/fable
Git Repo below 👇
An AI agent to make sense, needs many tools, or better to make many api calls, and imagine if you want to subscribe to each of em and never never to their cap! How come you make that agent into a profitable business? The solution is a proxy that lets you pay just for the API calls you make! Here @treg_ai is doing that and more: tool routing, team account, remvoe the need to share api keys between team members and many more.
Introducing Claude for people-search
No more $600 subscriptions, juts $0.0089/lead
Claude is now the best people finder in the world
#1 on People Search Bench
Fully open source, 0% markup
Try it at treg.to/people-search
Git Repo below 👇
Ok the mystery the ox-alpha is GLM-5.3-Flash, And modelprint.ai was right! ox alpha was the first real test I did. day one, the fingerprint said glm family, 4 of 4 tokenizer probes. day five, z.ai confirmed it. today i reran it on the official glm-5.3-flash endpoint: same values, 47, 56, 60, 90. the label changed, the model did not.
anyway the guessing game was fun, but it was not the point for me. but the real question: can you fingerprint a model from its behavior alone, without access to weights? even for open models, reading the weights is slow. behavior is fast.
if that works, you get something bigger: a version control system for models that does not depend on the lab, the provider, or anyone's word. the API itself is the source of truth.
so now it is a tool, and it is open. join, contribute, add probes. and when the next mystery model appears, we run it again. modelprint.ai
Introducing GLM-5.3-Flash
- Leading capabilities at a highly competitive price
- Natively multimodal with a 1M-token context window
- A 320B-A18B model released under the MIT License
- Previously previewed as Ox Alpha, running entirely on Chinese AI chips
Blog: z.ai/blog/glm-5.3-flash
Available now across all official platforms:
Weights: huggingface.co/zai-org/GLM-5…
API: docs.z.ai/guides/llm/glm-5.3…
Coding Plan: z.ai/subscribe
ZCode: zcode.z.ai/en
Chat: chat.z.ai
AutoClaw: autoclaw.z.ai
Bloomberg says Z.ai confirmed it: Ox Alpha is GLM. I did not guess. 5 days ago I fingerprinted the API: the GLM family matched on all 4 tokenizer probes. MiMo 2 of 4. Gemini 1 of 9. Grok and Claude 0. The measurement called it before the label. modelprint.ai
update: the official name is out, glm-5.3-flash. i reran the fingerprint on the new endpoint and compared with day one of the stealth run: tokenizer probes 47, 56, 60, 90 then. 47, 56, 60, 90 now. same template delta. the label changed, the model did not. modelprint.ai/history
The fingerprint from launch day, still open:
xitter.cf/unclecode/status/20907…
Bloomberg's confirmation: xitter.cf/business/status/209254…
Anyone can re-run it. No leaks, no sources, just the API telling on itself.
The new AI model that’s swept to the top of online usage charts with high performance at zero cost was created by China’s Z.AI, also known as Zhipu bloomberg.com/news/articles/…
Back to OxAlpha fingerpting, the modelprint has a home now: modelprint.ai And a memory. Every day it fingerprints 21 models, each locked to one provider, and records what it finds. A new entry is written only when a value actually changes. modelprint.ai/history So when someone says a model "feels different today", there is now a place to check instead of guess.
Everyone guessed who made Ox Alpha today. I fingerprinted it instead. One page, 9 infrastructure probes: tokenizers, error codes, hidden templates. I ran • the mystery model against 12 suspects. One family matches every tokenizer test: GLM. Plumbing does not lie. Tool in reply, bring your own keys. Link in reply
Everyone guessed who made Ox Alpha today. I fingerprinted it instead. One page, 9 infrastructure probes: tokenizers, error codes, hidden templates. I ran • the mystery model against 12 suspects. One family matches every tokenizer test: GLM. Plumbing does not lie. Tool in reply, bring your own keys. Link in reply
Update: modelprint now has free credits, $100 shared between everyone daily. Fingerprint any model with zero keys: xitter.cf/unclecode/status/20911…
Update: I put $100 of API credits into modelprint, shared between everyone, refreshed daily. You can now fingerprint a model with zero keys, zero setup. One click.
The plan is bigger than one mystery model. The model behind an API name can change without notice. I want fingerprinting to be the standard check: developers know what they really call, businesses verify what they pay for.
Update: I put $100 of API credits into modelprint, shared between everyone, refreshed daily. You can now fingerprint a model with zero keys, zero setup. One click.
The plan is bigger than one mystery model. The model behind an API name can change without notice. I want fingerprinting to be the standard check: developers know what they really call, businesses verify what they pay for.
Everyone guessed who made Ox Alpha today. I fingerprinted it instead. One page, 9 infrastructure probes: tokenizers, error codes, hidden templates. I ran • the mystery model against 12 suspects. One family matches every tokenizer test: GLM. Plumbing does not lie. Tool in reply, bring your own keys. Link in reply
Try it: unclecode.github.io/modelpri…
Free credits cover every model under $1.50/M input, including glm-5.3. Your own key removes all limits, and the :free models cost $0. Every probe is a one-file plugin, contributions open: github.com/unclecode/modelpr…
Unclecode (Hossein) retweeted
大神出手了，以后新模型是哪家的不用猜了。
这个modelprint开源工具解决了这个问题，今天的Ox Alpha 确定了是GLM 新模型！
工具地址：
unclecode.github.io/modelpri…
开源代码库在这里：
github.com/unclecode/modelpr…
Everyone guessed who made Ox Alpha today. I fingerprinted it instead. One page, 9 infrastructure probes: tokenizers, error codes, hidden templates. I ran • the mystery model against 12 suspects. One family matches every tokenizer test: GLM. Plumbing does not lie. Tool in reply, bring your own keys. Link in reply
Update: tested the last untested suspect through Xiaomi's official MiMo API. mimo-v2.5 matches the mystery on 2 of 4 tokenizer probes, v2.5-pro on 0 v2.5-pro on 0 of 4. of 4. The emoji probe separates hardest: 90 tokens vs 57. GLM stays the only 4/4. Last suspect cleared.
Run it: unclecode.github.io/modelpri…
No install. Your API keys stay in your browser tab, calls go straight to the provider. Every probe is one small JS file, and adding your own is a pull request: github.com/unclecode/modelpr… start it, we can add more probs for future models