Model comparison
Gemma-4-31B-JANG_4M-CRACK vs o3
Capability radar, value bars, and side-by-side posture.
Gemma-4-31B-JANG_4M-CRACK
Not scored
value score
o3
93
value score
Gemma-4-31B-JANG_4M-CRACK ctx
—
o3 ctx
200K
Capability radar
Value · context · multimodal · openness · speed posture
Value head-to-head
- o393
Attribute tape
| Field | Gemma-4-31B-JANG_4M-CRACK | o3 |
|---|---|---|
| Developer | DealignAI | OpenAI |
| Context | — | 200K |
| Modalities | Text, Code | |
| Openness | — | Proprietary Model |
| Speed | — | — |
| Price | — | $10/1M in · $40/1M out |
| Value score | — | 93 |
| Summary | An abliterated (safety-removed) 31B-parameter multimodal variant of Google's Gemma 4, formatted for JANG v2 and optimized for local MLX inference on Apple Silicon. Intended for unrestricted and security-research-style prompting rather than standard aligned assistant use. | OpenAI reasoning model focused on hard problems. |