|
Download README.md from stealth-model/iceman-contributor: direct link, hf CLI and curl.
- Browser
- Download file 2.62 kB
-
https://huggingface.co/stealth-model/iceman-contributor/resolve/main/README.md
- Command line
-
hf download hf://stealth-model/iceman-contributor/README.md
-
curl -L -o README.md https://huggingface.co/stealth-model/iceman-contributor/resolve/main/README.md
2.62 kB
| license: cc-by-4.0 | |
| tags: | |
| - iceman-contributor | |
| - stealth-model | |
| - stealthmark | |
| - coding | |
| - reasoning | |
| - long-context | |
| - tool-use | |
| thumbnail: https://stealthmodels.com/iceman-contributor/assets/social-card.png | |
| <img src="https://huggingface.co/stealth-model/iceman-contributor/resolve/main/panels/hero.webp" alt="A boy made of cracked ice lunging beside the name Iceman Contributor, 417 billion tokens on day one, two benchmark results and a 23 of 23 tokenizer match."> | |
| **Iceman Contributor counts tokens exactly like MAI Thinking 1 and Glyph Cluster, on 23 of 23 phrases, and no company has claimed it yet.** | |
| By [Stealth Models](https://stealthmodels.com/). | |
| <img src="https://huggingface.co/stealth-model/iceman-contributor/resolve/main/panels/verdict.webp" alt="StealthMark 1.8.10 results for Iceman Contributor: benchmark score 19.2, General Intelligence 11.7, Visual Intelligence 31.0. Strongest at composing a whole scene: 46.4 on Beach."> | |
| <img src="https://huggingface.co/stealth-model/iceman-contributor/resolve/main/panels/results.webp" alt="Every StealthMark result of Iceman Contributor: Beach 46.4, Selfie 39.3, Raccoon 29.8, Ground Truth 20.2, Mona Lisa 19.3, Newton's Mirror 10.1, Latch 3.6, Foresight 3.2."> | |
| [Compare it with Glyph Cluster](https://stealthmodels.com/compare/glyph-cluster-vs-iceman-contributor) | |
| <img src="https://huggingface.co/stealth-model/iceman-contributor/resolve/main/panels/case.webp" alt="The case file. First seen 24 Oct 2026, revealed not yet, developer unknown, context 262,144 tokens, input Text, tool calls Yes, speed in our tests about 46 tok/s. 24 Oct 2026: Appears without an announcement: about 417 billion tokens on its first full day. 24 Oct 2026: A post on X matches its token counts to Microsoft’s MAI Thinking 1. 25 Oct 2026, our finding: Our test: the same tokenizer family as Glyph Cluster, our MAI lead."> | |
| [See the token-count test](https://stealthmodels.com/iceman-contributor#tokenizer) | |
| <details><summary>Facts as text</summary> | |
| | Fact | Value | | |
| |---|---| | |
| | First seen | 24 Oct 2026 | | |
| | Revealed | not yet | | |
| | Developer | unknown | | |
| | Context | 262,144 tokens | | |
| | Input | Text | | |
| | Tool calls | Yes | | |
| | Speed in our tests | about 46 tok/s | | |
| | Access | Active; prompts may be used for training | | |
| </details> | |
| ## Watch | |
| [](https://www.youtube.com/watch?v=l5hdM8lsPfI) | |
| --- | |
| [StealthMark leaderboard](https://huggingface.co/spaces/stealth-model/stealthmark)<br> | |
| Text and images: CC BY 4.0, stealthmodels.com. | |