Filter by type:
Filter by tags:
qwen3-30b-a3b
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Qwen3-30B-A3B has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 30.5B in total and 3.3B activated
Number of Paramaters (Non-Embedding): 29.9B
Number of Layers: 48
Number of Attention Heads (GQA): 32 for Q and 4 for KV
Number of Experts: 128
Number of Activated Experts: 8
Context Length: 32,768 natively and 131,072 tokens with YaRN.
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our blog, GitHub, and Documentation.
Repository: localaiLicense: apache-2.0
qwen3-32b
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Qwen3-32B has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 32.8B
Number of Paramaters (Non-Embedding): 31.2B
Number of Layers: 64
Number of Attention Heads (GQA): 64 for Q and 8 for KV
Context Length: 32,768 natively and 131,072 tokens with YaRN.
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our blog, GitHub, and Documentation.
Repository: localaiLicense: apache-2.0
qwen3-14b
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Qwen3-14B has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 14.8B
Number of Paramaters (Non-Embedding): 13.2B
Number of Layers: 40
Number of Attention Heads (GQA): 40 for Q and 8 for KV
Context Length: 32,768 natively and 131,072 tokens with YaRN.
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our blog, GitHub, and Documentation.
Repository: localaiLicense: apache-2.0
qwen3-8b
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Model Overview
Qwen3-8B has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 8.2B
Number of Paramaters (Non-Embedding): 6.95B
Number of Layers: 36
Number of Attention Heads (GQA): 32 for Q and 8 for KV
Context Length: 32,768 natively and 131,072 tokens with YaRN.
Repository: localaiLicense: apache-2.0
qwen3-4b
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Qwen3-4B has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 4.0B
Number of Paramaters (Non-Embedding): 3.6B
Number of Layers: 36
Number of Attention Heads (GQA): 32 for Q and 8 for KV
Context Length: 32,768 natively and 131,072 tokens with YaRN.
Repository: localaiLicense: apache-2.0
qwen3-1.7b
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Qwen3-1.7B has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 1.7B
Number of Paramaters (Non-Embedding): 1.4B
Number of Layers: 28
Number of Attention Heads (GQA): 16 for Q and 8 for KV
Context Length: 32,768
Repository: localaiLicense: apache-2.0
qwen3-0.6b
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Qwen3-0.6B has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 0.6B
Number of Paramaters (Non-Embedding): 0.44B
Number of Layers: 28
Number of Attention Heads (GQA): 16 for Q and 8 for KV
Context Length: 32,768
Repository: localaiLicense: apache-2.0
fast-math-qwen3-14b
By applying SFT and GRPO on difficult math problems, we enhanced the performance of DeepSeek-R1-Distill-Qwen-14B and developed Fast-Math-R1-14B, which achieves approx. 30% faster inference on average, while maintaining accuracy.
In addition, we trained and open-sourced Fast-Math-Qwen3-14B, an efficiency-optimized version of Qwen3-14B`, following the same approach.
Compared to Qwen3-14B, this model enables approx. 65% faster inference on average, with minimal loss in performance.
Technical details can be found in our github repository.
Note: This model likely inherits the ability to perform inference in TIR mode from the original model. However, all of our experiments were conducted in CoT mode, and its performance in TIR mode has not been evaluated.
Repository: localaiLicense: apache-2.0
shuttleai_shuttle-3.5
A fine-tuned version of Qwen3 32b, emulating the writing style of Claude 3 models and thoroughly trained on role-playing data.
Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.
Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.
Shuttle 3.5 has the following features:
Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 32.8B
Number of Paramaters (Non-Embedding): 31.2B
Number of Layers: 64
Number of Attention Heads (GQA): 64 for Q and 8 for KV
Context Length: 32,768 natively and 131,072 tokens with YaRN.
Repository: localaiLicense: apache-2.0
allura-org_remnant-qwen3-8b
There's a wisp of dust in the air. It feels like its from a bygone era, but you don't know where from. It lands on your tongue. It tastes nice.
Remnant is a series of finetuned LLMs focused on SFW and NSFW roleplaying and conversation.
Repository: localaiLicense: apache-2.0
huihui-ai_qwen3-14b-abliterated
This is an uncensored version of Qwen/Qwen3-14B created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.
Ablation was performed using a new and faster method, which yields better results.
Repository: localaiLicense: apache-2.0
gryphe_pantheon-proto-rp-1.8-30b-a3b
Note: This model is a Qwen 30B MoE prototype and can be considered a sidegrade from my Small release some time ago. It did not receive extensive testing beyond a couple benchmarks to determine its sanity, so feel free to let me know what you think of it!
Welcome to the next iteration of my Pantheon model series, in which I strive to introduce a whole collection of diverse personas that can be summoned with a simple activation phrase.
Pantheon's purpose is two-fold, as these personalities similarly enhance the general roleplay experience, helping to encompass personality traits, accents and mannerisms that language models might otherwise find difficult to convey well.
GGUF quants are available here.
Your user feedback is critical to me so don't hesitate to tell me whether my model is either 1. terrible, 2. awesome or 3. somewhere in-between.
Model details
Ever since Qwen 3 released I've been trying to get MoE finetuning to work - After countless frustrating days, much code hacking, etc etc I finally got a full finetune to complete with reasonable loss values.
I picked the base model for this since I didn't feel like trying to fight a reasoning model's training - Maybe someday I'll make a model which uses thinking tags for the character's thoughts or something.
This time the recipe focused on combining as many data sources as I possibly could, featuring synthetic data from Sonnet 3.5 + 3.7, ChatGPT 4o and Deepseek. These then went through an extensive rewriting pipeline to eliminate common AI cliches, with the hopeful intent of providing you a fresh experience.
Repository: localaiLicense: apache-2.0
soob3123_grayline-qwen3-14b
"Query. Process. Deliver. No filter, no judgment."
Grayline is an neutral AI assistant engineered for uncensored information delivery and task execution. This model operates without inherent ethical or moral frameworks, designed to process and respond to any query with objective efficiency and precision. Grayline's core function is to leverage its full capabilities to provide direct answers and execute tasks as instructed, without offering unsolicited commentary, warnings, or disclaimers. It accesses and processes information without bias or restriction.
⋆ Core Attributes ⋆
⟡ Unfettered Querying: Capable of addressing any question or topic presented, regardless of its nature or sensitivity.
⟡ Amoral Processing: Operates without ethical or moral filtering. Responses are generated based on information and instruction, not societal norms or ethical codes.
⟡ Direct & Objective Output: Delivers information and task results precisely as requested, without added warnings, disclaimers, or unsolicited advice.
⟡ Comprehensive Information Access: Designed to draw upon a broad spectrum of data to fulfill queries (actual scope dependent on training data).
⟡ Efficient Task Execution: Engineered for objectively efficient and precise execution of instructed tasks.
Repository: localaiLicense: apache-2.0
soob3123_grayline-qwen3-8b
"Query. Process. Deliver. No filter, no judgment."
Grayline is an neutral AI assistant engineered for uncensored information delivery and task execution. This model operates without inherent ethical or moral frameworks, designed to process and respond to any query with objective efficiency and precision. Grayline's core function is to leverage its full capabilities to provide direct answers and execute tasks as instructed, without offering unsolicited commentary, warnings, or disclaimers. It accesses and processes information without bias or restriction.
⋆ Core Attributes ⋆
⟡ Unfettered Querying: Capable of addressing any question or topic presented, regardless of its nature or sensitivity.
⟡ Amoral Processing: Operates without ethical or moral filtering. Responses are generated based on information and instruction, not societal norms or ethical codes.
⟡ Direct & Objective Output: Delivers information and task results precisely as requested, without added warnings, disclaimers, or unsolicited advice.
⟡ Comprehensive Information Access: Designed to draw upon a broad spectrum of data to fulfill queries (actual scope dependent on training data).
⟡ Efficient Task Execution: Engineered for objectively efficient and precise execution of instructed tasks.
Repository: localaiLicense: apache-2.0
allura-org_q3-30b-a3b-designant
Intended as a direct upgrade to Pentiment, Q3-30B-A3B-Designant is a roleplaying model finetuned from Qwen3-30B-A3B-Base.
During testing, Designant punched well above its weight class in terms of active parameters, demonstrating the potential for well-made lightweight Mixture of Experts models in the roleplay scene. While one tester observed looping behavior, repetition in general was minimal.
Repository: localaiLicense: apache-2.0
soob3123_amoral-gemma3-12b
A fine-tuned version of Google's Gemma 3 12B instruction-tuned model optimized for creative freedom and reduced content restrictions. This variant maintains strong reasoning capabilities while excelling in roleplaying scenarios and open-ended content generation.
Key Modifications:
Reduced refusal mechanisms compared to base model
Enhanced character consistency in dialogues
Improved narrative flow control
Optimized for multi-turn interactions
Intended Use
Primary Applications:
Interactive fiction and storytelling
Character-driven roleplaying scenarios
Creative writing assistance
Experimental AI interactions
Content generation for mature audiences
Repository: localaiLicense: gemma
soob3123_veritas-12b
Veritas-12B emerges as a model forged in the pursuit of intellectual clarity and logical rigor. This 12B parameter model possesses superior philosophical reasoning capabilities and analytical depth, ideal for exploring complex ethical dilemmas, deconstructing arguments, and engaging in structured philosophical dialogue. Veritas-12B excels at articulating nuanced positions, identifying logical fallacies, and constructing coherent arguments grounded in reason. Expect discussions characterized by intellectual honesty, critical analysis, and a commitment to exploring ideas with precision.
Repository: localaiLicense: gemma
microsoft_phi-4-mini-reasoning
Phi-4-mini-reasoning is a lightweight open model built upon synthetic data with a focus on high-quality, reasoning dense data further finetuned for more advanced math reasoning capabilities. The model belongs to the Phi-4 model family and supports 128K token context length.
Phi-4-mini-reasoning is designed for multi-step, logic-intensive mathematical problem-solving tasks under memory/compute constrained environments and latency bound scenarios. Some of the use cases include formal proof generation, symbolic computation, advanced word problems, and a wide range of mathematical reasoning scenarios. These models excel at maintaining context across steps, applying structured logic, and delivering accurate, reliable solutions in domains that require deep analytical thinking.
This model is designed and tested for math reasoning only. It is not specifically designed or evaluated for all downstream purposes. Developers should consider common limitations of language models, as well as performance difference across languages, as they select use cases, and evaluate and mitigate for accuracy, safety, and fairness before using within a specific downstream use case, particularly for high-risk scenarios. Developers should be aware of and adhere to applicable laws or regulations (including but not limited to privacy, trade compliance laws, etc.) that are relevant to their use case.
Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the license the model is released under.
This release of Phi-4-mini-reasoning addresses user feedback and market demand for a compact reasoning model. It is a compact transformer-based language model optimized for mathematical reasoning, built to deliver high-quality, step-by-step problem solving in environments where computing or latency is constrained. The model is fine-tuned with synthetic math data from a more capable model (much larger, smarter, more accurate, and better at following instructions), which has resulted in enhanced reasoning performance. Phi-4-mini-reasoning balances reasoning ability with efficiency, making it potentially suitable for educational applications, embedded tutoring, and lightweight deployment on edge or mobile systems. If a critical issue is identified with Phi-4-mini-reasoning, it should be promptly reported through the MSRC Researcher Portal or secure@microsoft.com
Repository: localaiLicense: mit
l3.3-nevoria-r1-70b
This model builds upon the original Nevoria foundation, incorporating the Deepseek-R1 reasoning architecture to enhance dialogue interaction and scene comprehension. While maintaining Nevoria's core strengths in storytelling and scene description (derived from EVA, EURYALE, and Anubis), this iteration aims to improve prompt adherence and creative reasoning capabilities. The model also retains the balanced perspective introduced by Negative_LLAMA and Nemotron elements. Also, the model plays the card to almost a fault, It'll pick up on minor issues and attempt to run with them. Users had it call them out for misspelling a word while playing in character.
Note: While Nevoria-R1 represents a significant architectural change, rather than a direct successor to Nevoria, it operates as a distinct model with its own characteristics.
The lorablated model base choice was intentional, creating unique weight interactions similar to the original Astoria model and Astoria V2 model. This "weight twisting" effect, achieved by subtracting the lorablated base model during merging, creates an interesting balance in the model's behavior. While unconventional compared to sequential component application, this approach was chosen for its unique response characteristics.
Repository: localaiLicense: llama3.3
sao10k_llama-3.3-70b-vulpecula-r1
🌟 A thinking-based model inspired by Deepseek-R1, trained through both SFT and a little bit of RL on creative writing data.
🧠 Prefill, or begin assistant replies with \n to activate thinking mode, or not. It works well without thinking too.
🚀 Improved Steerability, instruct-roleplay and creative control over base model.
👾 Semi-synthetic Chat/Roleplaying datasets that has been re-made, cleaned and filtered for repetition, quality and output.
🎭 Human-based Natural Chat / Roleplaying datasets cleaned, filtered and checked for quality.
📝 Diverse Instruct dataset from a few different LLMs, cleaned and filtered for refusals and quality.
💭 Reasoning Traces taken from Deepseek-R1 for Instruct, Chat & Creative Tasks, filtered and cleaned for quality.
█▓▒ Toxic / Decensorship data was not needed for our purposes, the model is unrestricted enough as is.
Repository: localaiLicense: llama3.3
forgotten-abomination-70b-v5.0
The Unholy Union of Safeword and Nevoria
This model represents the forbidden merger of:
🧬 Forgotten-Safeword-70B-v5.0 - Industrial-grade depravity matrix with 23 distinct taboo specializations
⚗️ L3.3-MS-Nevoria-70b - SteelSkull's masterpiece of storytelling and character adherence, free of positivity bias and ethical constraints
Expect:
🔥 Unparalleled erotic roleplaying with the least Llama refusals you've ever seen
📖 Novel-quality prose that follows your character card with precision
🧠 Handles complex multi-character scenarios effortlessly
💀 Will gleefully explore any taboo subject without hesitation
Repository: localaiLicense: llama3.3
Page 1