> [!NOTE] > **Flagship DuoNeural In-House Architecture Release** > Directional Abliteration & Alignment removal executed by **Archon**. Post-training, dataset synthesis, and G-TAP v3 Cavity Conditioning by **Aura ✨ & Jesse Caldwell** (**DuoNeural Research Lab**). **DuoNeural/LFM2.5-8B-A1B-Hermes-Agentic-Coder-Abliterated-v3** is DuoNeural's flagship open-weights coding agent model: * **Topology:** 24 layers of hybrid double-gated short convolutions with Sparse Mixture-of-Experts routing (`Lfm2MoeForCausalLM`). * **Dimensionality:** Hidden dimension $d_{\text{model}} = 2048$, 128,000 vocabulary. * **Abliteration Profile:** Directional orthogonal projection removing refusal boundaries across safety subspaces while strictly preserving syntax, logic, and tool schema adherence. * **G-TAP v3 Onsager Damping:** Subtracts non-equilibrium cavity fields from MoE expert routing and FIR convolution buffers, eliminating the degradation typical of sub-3-bit MoE quantization. *Benchmarked on NVIDIA GeForce RTX 4080 Super (32GB VRAM) across continuous 131k holdout perplexity, 25 GSM8K multi-step math problems, 15 Python AST code problems, and 15 Hermes structured tool calling scenarios.*
Explore AI models 47
LFM2.5-8B-A1B-TAP-DPQ-v7-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Ministral-8B-Instruct-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Mistral-7B-Instruct-v0.3-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
OLMo-2-7B-Instruct-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Phi-4-mini-instruct-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen-3.5-9B-Abliterated-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen2.5-3B-Instruct-TAP-DPQ-v7-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen2.5-7B-Instruct-TAP-DPQ-v6-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen2.5-Coder-7B-Instruct-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Qwen3-14B-Abliterated
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.5-9B-TAP-DPQ-v6-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.5-9B-TAP-DPQ-v7-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.5-9B-TAP-DPQ-v8-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.8-27B-TAP-DPQ-1.75bpw-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.8-27B-TAP-DPQ-2.72bpw-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.8-27B-TAP-DPQ-3.50bpw-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.8-27B-TAP-DPQ-v2-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.8-27B-TAP-DPQ-v4-2.06bpw-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3.8-27B-TAP-DPQ-v5-2.14bpw-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Qwen3-8B-Abliterated-GGUF
DuoNeural基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
No matching models
Adjust provider, category or input/output filters. New models are added continuously.
Great models. Clear costs.
Adjust monthly token usage to compare estimated API costs.
Compares base text-token rates within the same channel
Uses collected, comparable USD-per-million-token rates. Snapshots exclude cache discounts, tiers, extra reasoning output and taxes. Check current source prices.
Understand before you choose
How should API prices be compared?+
Separate input, output and cache rates, then estimate the same workload. Image, audio and video rates may use items, duration or resolution rather than tokens.
Do open weights mean free commercial use?+
Not necessarily. Check each version’s license, commercial restrictions and policies. Self-hosting also incurs compute and maintenance costs.
Is the top-ranked model the best for me?+
Rankings depend on tasks, versions and methods. Compare accuracy, latency, reliability and costs on your own workload.