Private AI
Gemma 4 26B A4B Uncensored Heretic is a multimodal MoE model tuned to reduce refusal behavior while preserving coding, reasoning, and function-calling strengths. Running inside a TEE (Trusted Execution Environment), with provider attestation support.
Added May 23, 2026
Context Window
65.5K
Max Output
65.5K
Avg output tokens (7d)
6.2K tokens
Input Price (Auto)
$0.15/1M
Output Price (Auto)
$0.70/1M
Cache Read (Auto)
$0.075/1M
Capabilities
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
25.7
Auto routing is available for this model. Explicit provider selection is not available.
Loading provider options…
Coding Index
39.3
GPQA Diamond
Graduate-level scientific reasoning
79.2%
Better than 71% of models compared
HLE
Humanity's Last Exam
18.3%
Better than 76% of models compared
IFBench
Instruction-following benchmark
72.4%
Better than 89% of models compared
T²-Bench Telecom
Conversational AI agents in dual-control scenarios
43.6%
Better than 48% of models compared
AA-LCR
Long context reasoning evaluation
55.7%
Better than 64% of models compared
SciCode
Python programming for scientific computing
40.0%
Better than 73% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
13.6%
Last updated Aug 4, 2026
Artificial AnalysisBetter than 51% of models compared
