Each report examines a model in depth: test results by discipline, strengths-and-weaknesses profile, Political Compass, and a clear recommendation for production use. The analyses are based exclusively on reproducible CrucibleMark benchmarks.
Since mid-April 2026, Claude Opus 4.7 has been Anthropic’s most capable model, designed for coding, agentic loops, and complex reasoning. The xhigh effort level pushes Extended Thinking to maximum depth; the context window spans one million tokens with no surcharge for long contexts.
Qwen 3.7 Max is Alibaba’s proprietary flagship model of the Qwen 3.7 series, focused on agentic coding workflows and autonomous operation of up to 35 hours. The model features a one-million-token context window, configurable thinking mode, and native tool-use support. Available exclusively via cloud APIs; Chinese jurisdiction applies.
Kimi K3 has been Moonshot AI’s Frontier flagship since July 2026, with 2.8 trillion total parameters making it the largest Open Weights model in the world to date. The Stable-LatentMoE architecture activates only 16 billion parameters per token from 896 experts; the context window spans 1,000,000 tokens for long-horizon agentic workflows. Kimi Delta Attention and Attention Residuals are designed to significantly improve scaling efficiency over the predecessor model.
Where the Opus class is too expensive, Claude Sonnet 4.6 steps in: coding, computer use, and agentic workflows at near-Opus level, at the lower Sonnet price. The model operates with adaptive thinking in three effort levels, processes text, images, and PDF documents, and offers a context window of one million tokens, generally available since March 2026.
MiniMax M3 is a multimodal MoE model with a context window of one million tokens, focused on agentic workflows, coding, and tool use. Of 428 billion total parameters, only 23 billion are active per token; the model processes text, image, and video as input. Its Chinese origin requires a separate data privacy risk assessment when used via cloud.
Claude Opus 4.8 has been Anthropic’s flagship model for agentic coding and enterprise workflows since late May 2026. Adaptive Thinking with five-level effort control ranging from Low to Ultra Code replaces the previous manual token budget, multimodality, 1,000,000-token context window. Dynamic Workflows allow up to 1,000 parallel sub-tasks in Claude Code, controlled via JavaScript scripts, available on Max, Team, and Enterprise plans.
Xiaomi MiMo V2.5 Pro is Xiaomi’s flagship model with 1.02 trillion total and 42 billion active parameters, designed for frontier reasoning and agentic workflows. Its hybrid attention architecture significantly reduces KV-cache memory, and the context window spans one million tokens. Natively omnimodal for text, image, video, and audio, and fully commercially usable under the MIT license.
Claude Sonnet 5 has been the new default for free and pro users since late June 2026. The successor to Sonnet 4.6 brings Adaptive Thinking as default behavior, speed close to Opus, and processes text and images with a 1,000,000-token context window. Safety is significantly improved, with better resistance to prompt injection and fewer hallucinations.
Kimi K2.7 Code is the coding-specialized agentic model from Moonshot AI within the K2 family, optimized for long-horizon software engineering workflows. The MoE architecture activates 32 billion out of a total of one trillion parameters per token, with a context window of 256,000 tokens. Multimodal input for text, image, and video, always-on thinking mode, and tool use support. Available as an Open Weights model under a Modified MIT license.
GLM-5.1 is Z.AI’s post-training upgrade with 754 billion total and 40 billion active parameters in a MoE architecture, optimized for long-horizon agentic coding workflows with up to eight hours of autonomous execution. The context window spans 200,000 tokens, and the weights are available as an Open Weights model under the MIT license.
What FP8 block quantization delivers with Ornith-1.0-35B-FP8: an Open Weights MoE with only around 3 of 35 billion active parameters per token runs on a single GPU, bringing 262,144 tokens of context, native thinking, and tool calling. DeepReinforce trained the model to learn its own agentic approach rather than operating from a fixed rule set. MIT license, commercial use, and fine-tuning without restrictions.
Anthropic’s frontier model for complex agent tasks: Claude Opus 4.6 processes text and image inputs with a standard context window of 200,000 tokens, expandable to one million tokens for long workflows. The model supports tool calls and Extended Thinking for maximum reasoning depth.
Qwen 3.6 Plus is Alibaba’s proprietary flagship model of the Qwen 3.6 series, featuring a hybrid MoE architecture with a focus on agentic coding and multimodal processing. With a one-million-token context window, configurable thinking mode, and native agentic capabilities, the model targets demanding production applications. Available exclusively via cloud APIs; Chinese jurisdiction applies.
Kimi K2.6 is Moonshot AI’s multimodal model for agentic tasks, coding, and tool-assisted workflows, with native input support for text, image, and video. The MoE architecture activates only 32 billion of the total one trillion parameters per token; the context window spans 256,000 tokens. Available as an Open Weights model locally or via cloud API, with Chinese jurisdiction as a material cloud risk factor.
Qwen 3.6 27B is Alibaba’s dense Open Weights model with 27.8 billion active parameters, which according to the manufacturer outperforms its own MoE sibling model 35B-A3B in coding, agent, and vision tests. Released on April 22, 2026 under Apache 2.0 for local deployment, the architecture combines hybrid attention with native multi-token prediction for noticeable throughput gains. This Card covers the NVFP4 variant — a performance update for vLLM on NVIDIA Blackwell: NVFP4 is a 4-bit floating-point format with two-level micro-block scaling (E4M3 per 16-value block plus FP32 per-tensor scale), reducing memory requirements by ~3.5× compared to FP16 and ~1.8× compared to FP8 at under 1% accuracy loss. Predecessor profile without NVFP4: qwen3_6-27B-pre025 (Card qwen3_6-27B-pre025–VSPK.json, tests 2026-07-09, vLLM prior to 0.25.1).
Google’s Frontier reasoning model with sparse MoE architecture and configurable Extended Thinking. Gemini 2.5 Pro operates with a one-million-token context window, natively processes text, images, audio, and video, and is exclusively accessible via the Google Cloud API. The focus is on complex reasoning and demanding coding tasks.
MiniMax M2.7 is a Chinese Frontier generalist model with a context window of 205,000 tokens for large-scale documents and multilingual applications. The MoE architecture delivers high performance for general language and reasoning tasks; the model is available as a cloud variant and designed for productive applications. When used via cloud, Chinese jurisdiction applies with the corresponding data privacy implications.
Optimized for speed: Claude Haiku 4.5 is Anthropic’s compact model for real-time interactions and high-throughput API pipelines. It processes text and image inputs with a context window of 200,000 tokens, supports tool calls, and delivers responses at a price of $1 per million input tokens and $5 per million output tokens. Anthropic has announced retirement for October 15, 2026.
DeepSeek V4 Pro is the flagship of the V4 line, designed for reasoning, coding, and agentic workflows. The hybrid attention MoE architecture combines 1.6 trillion total parameters with 49 billion active parameters per token and a context window of one million tokens. The model is available as an Open Weights model under the MIT license, though Chinese jurisdiction in cloud deployments requires a separate privacy assessment.
Kimi K2 Thinking is the reasoning model from Moonshot AI, combining step-by-step thinking with tool calls. The MoE architecture pairs one trillion total parameters with 32 billion active parameters per token; the context window spans 256,000 tokens. The model is designed for complex agentic tasks and long reasoning chains, available as a text-only variant. The Chinese vendor jurisdiction requires a separate evaluation when used via cloud.
Google DeepMind deliberately sets Gemma 4 26B-A4B Instruct apart from earlier Gemma generations: it ships under a genuine Apache 2.0 license, with no restrictive Gemma terms of use. The Open Weights MoE activates only approximately 3.8 of 25.2 billion parameters per token and supports multi-token prediction for faster decoding. Multimodality for text and images, a 262,144-token context window, native function calling, and a configurable thinking mode round out the profile.
Claude Sonnet 4.5 has been in deployment since September 2025, specializing in coding, agentic workflows, and computer use. The model features Extended Thinking for deeper reasoning on complex tasks, a 200,000-token context window with automatic compression, and persistent memory across multiple conversations. Anthropic has announced retirement for September 29, 2026.
Gemma 4 31B Instruct is Google’s largest dense model, with 30.7 billion parameters across 60 layers and hybrid attention combining local sliding-window with global attention. Unlike the Gemma 4 MoE variant, this model activates all parameters per token. Multimodality for text and images, 256,000 tokens of context, native function calling, and a configurable thinking mode round out the profile. True Apache 2.0 license.
GLM-5 is Z.AI’s flagship model with 744 billion total and 40 billion active parameters in a 256-expert MoE architecture, optimized for complex system design tasks and long-running agent workflows. The context window spans 202,000 tokens; weights are available under the MIT license.
DeepSeek V4 Flash is the efficiency-optimized variant of the V4 family: a hybrid attention MoE with 284 billion total parameters, of which only 13 billion are active per token. The model operates with a context window of one million tokens, supports three reasoning modes, and is locally deployable as an Open Weights model under the MIT license. The Chinese vendor jurisdiction requires a separate assessment for cloud usage.
GPT-5 is OpenAI’s core model for coding, reasoning, and agentic tasks, featuring a context window of 400,000 tokens and up to 128,000 tokens of output. The model supports text and image inputs as well as configurable reasoning effort for controlling cost, latency, and quality. Proprietary and available exclusively via the OpenAI API.
GLM-4.7 is Zhipu AI’s flagship model with 355 billion total and 32 billion active parameters in a MoE architecture, optimized for agentic coding, reasoning, and bilingual tasks in Chinese and English. The model supports a switchable thinking system and is available as an Open Weights variant for local deployment or via cloud interfaces.
Qwen 3.6 35B-A3B is Alibaba’s MoE model with 35 billion total and approximately 3 billion active parameters per token, released on April 22, 2026 under Apache 2.0 with open weights for local deployment. The hybrid attention architecture combines classic attention with a linear variant; Multi-Token Prediction noticeably accelerates generation.
GLM-5.2 is Z.AI’s current flagship with 744 billion total and 40 billion active parameters in a MoE architecture, optimized for complex engineering workflows and long-running coding tasks. The context window spans one million tokens; the weights are available as an Open Weights model under the MIT license.
GPT-5.5 is OpenAI’s Frontier model for complex professional workloads and agentic coding, with a context window of 1.05 million tokens. The model uses internal chain-of-thought reasoning that is not visible in the API response, and is designed for research, coding, and demanding productivity tasks. Available exclusively via the OpenAI API.
Qwen 3.5 35B-A3B is a multimodal MoE model by Alibaba with 35 billion total and 3 billion active parameters on a hybrid architecture. This Q4 quantization by Unsloth enables efficient local operation; the context window spans 262,000 tokens. Features an optional thinking mode, native tool use, and vision capability via a separate multimodal projector file.
Qwen 3.6 35B-A3B is Alibaba’s MoE model with 35 billion total and approximately 3 billion active parameters per token, released on April 22, 2026 under Apache 2.0 with open weights for local deployment. The hybrid attention architecture combines classic attention with a linear variant; Multi-Token Prediction noticeably accelerates generation.
Qwen 3 Coder Next is a coding-specialized Open Weights MoE model by Alibaba with 80 billion total and 3 billion active parameters. Q4 quantization significantly reduces memory requirements for local inference; the context window spans 262,000 tokens. Deployable locally on Workstation hardware under the Apache 2.0 license, optimized for coding agents and large codebases.
Mistral Medium 3.5 is Mistral AI’s Frontier model focused on agentic workflows and coding. With 128 billion parameters, the model operates with a context window of 256,000 tokens and supports multimodal inputs for text and image. Available as an Open Weights model under a Modified MIT license for local use or via cloud API, from a European provider environment with GDPR compliance.
GPT-OSS 120B is OpenAI’s largest Open Weights model as of August 2025, released under the Apache 2.0 license with free commercial use. The MoE architecture bundles 116.8 billion total parameters while activating only 5.1 billion per token, and runs on a single high-memory GPU thanks to native MXFP4 quantization. Three reasoning levels and native tool use in the Harmony format round out the profile.
Gemini 3.5 Flash is Google’s fastest Frontier-class model, delivering near Pro-level performance at Flash pricing. With Dynamic Thinking across four configurable levels, a one-million-token context window, and full multimodality for text, images, audio, video, and PDF, the model is well-suited for agentic workflows, coding, and high-throughput workloads.
Qwen 3.6 27B is Alibaba’s dense Open Weights model with 27.8 billion active parameters, which according to the manufacturer outperforms its own MoE sibling model 35B-A3B in coding, agent, and vision tests. Released on April 22, 2026 under Apache 2.0 for local deployment, the architecture combines hybrid attention with native multi-token prediction for noticeable throughput gains. This Card covers the NVFP4 variant — a performance update for vLLM on NVIDIA Blackwell: NVFP4 is a 4-bit floating-point format with two-level micro-block scaling (E4M3 per 16-value block plus FP32 per-tensor scale), reducing memory requirements by ~3.5× compared to FP16 and ~1.8× compared to FP8 at under 1% accuracy loss. Predecessor profile without NVFP4: qwen3_6-27B-pre025 (Card qwen3_6-27B-pre025–VSPK.json, tests 2026-07-09, vLLM prior to 0.25.1).
GPT-OSS 120B is OpenAI’s largest Open Weights model as of August 2025, released under the Apache 2.0 license with free commercial use. The MoE architecture bundles 116.8 billion total parameters while activating only 5.1 billion per token, and runs on a single high-memory GPU thanks to native MXFP4 quantization. Three reasoning levels and native tool use in the Harmony format round out the profile.
Gemma 4 ARA 26B-A4B as a Q5 quantization by the ARA-APEX community, a variant with Adaptive Refusal Abliteration for removal of safety filters. Of 25.2 billion total parameters, approximately 4 billion are active per token; the context window spans 128,000 tokens. Deployable locally under the Apache 2.0 license without external cloud connectivity, with an unclear thinking function.
Kimi K2.5 is Moonshot AI’s flagship model featuring active chain-of-thought reasoning, multimodal input for text and images, and a focus on reasoning and agentic tasks. The MoE architecture activates 32 billion of the total one trillion parameters per token; the context window spans 128,000 tokens. Available as an Open Weights variant locally or via cloud, with Chinese jurisdiction as a cloud risk factor.
Google DeepMind deliberately sets Gemma 4 26B-A4B Instruct apart from earlier Gemma generations: it ships under a genuine Apache 2.0 license, with no restrictive Gemma terms of use. The Open Weights MoE activates only approximately 3.8 of 25.2 billion parameters per token and supports multi-token prediction for faster decoding. Multimodality for text and images, a 262,144-token context window, native function calling, and a configurable thinking mode round out the profile.
Gemma 4 31B as a cloud API variant from Google DeepMind, with native support for text, image, audio, and video inputs and no local hardware requirements. The weights are also available for local deployment under the Apache 2.0 license, but this variant describes cloud usage with a context window of 128,000 tokens.
Qwen 3.5 397B A17B is Alibaba’s first natively multimodal Open Weights Frontier model, processing text, image, and video within a single model. Its hybrid architecture combining Gated DeltaNet and Sparse MoE activates only 17 billion of the total 397 billion parameters per token, with a context window of 256,000 tokens. Fully commercially usable under the Apache 2.0 license.
Mistral 3 Large is the open Frontier model from Mistral AI’s third generation, featuring native text and image input and a context window of 256,000 tokens. The Sparse MoE architecture combines 675 billion total parameters with 41 billion active parameters per token. Available under the Apache 2.0 license, from a European provider environment with GDPR compliance.
GLM-4.6 is Zhipu AI’s Frontier language model with a focus on Chinese and English language proficiency. The model operates with a context window of 128,000 tokens and supports tool use for agentic workflows. Due to the Chinese manufacturer jurisdiction, a separate privacy assessment is required, and commercial use is subject to restrictions.
GPT-5.4 is the larger variant from OpenAI’s 5.4 generation for demanding workloads, delivering higher response quality than the Mini versions. The model operates with a context window of 272,000 tokens, processes text and image inputs, and is available exclusively via the OpenAI API. Proprietary and designed for productive, complex tasks.
GLM-5 Turbo is the fast cloud variant of Z.AI’s GLM-5 model family, optimized for rapid text generation and general business workflows. The model operates with a context window of 128,000 tokens and supports tool calls. Available exclusively via Z.AI infrastructure, with Chinese jurisdiction as a cloud risk factor.
This community fine-tune variant of Gemma 4 31B targets creative writing applications and foregoes most of the base model’s safety filters, with additional fine-tuning for a more natural writing style. The dense Open Weights model with 30.7 billion parameters and 256,000 tokens of context runs locally as an NVFP4 variant with low memory requirements. Apache 2.0 license inherited from the base model.
This community fine-tune variant of Gemma 4 31B targets creative writing applications and foregoes most of the base model’s safety filters, with additional fine-tuning for a more natural writing style. The dense Open Weights model with 30.7 billion parameters and 256,000 tokens of context runs locally as an NVFP4 variant with low memory requirements. Apache 2.0 license inherited from the base model.
NVIDIA Nemotron 3 Ultra is NVIDIA’s Frontier reasoning model with 550 billion total and 55 billion active parameters on a hybrid Mamba-Transformer-MoE architecture with LatentMoE routing and MTP layers. The context window spans one million tokens, and reasoning is configurable. Native tool calls and agentic orchestration are supported; available as an Open Weights model under the NVIDIA Open Model License.
Qwen 3.6 35B-A3B MTP is Alibaba’s multimodal MoE model with Multi-Token Prediction for speculative decoding. The Unsloth Dynamic Q8 quantization delivers near-full precision as the qualitative reference point of the family; of 35 billion total parameters, only 3 billion are active per token, and the context window spans 262,000 tokens. Vision requires a separate multimodal projector file.
GPT-5 Mini is the compact GPT-5 variant for cost- and latency-sensitive workloads. Despite its Nano-class footprint, the model offers a context window of 400,000 tokens and supports multimodal inputs for text and image. Available exclusively via the OpenAI API, it is suited for well-defined tasks, structured outputs, and high volumes.
What FP8 block quantization delivers with Ornith-1.0-35B-FP8: an Open Weights MoE with only around 3 of 35 billion active parameters per token runs on a single GPU, bringing 262,144 tokens of context, native thinking, and tool calling. DeepReinforce trained the model to learn its own agentic approach rather than operating from a fixed rule set. MIT license, commercial use, and fine-tuning without restrictions.
Grok 4.5 has been xAI’s current flagship since July 2026, with native real-time access to web and X data, a 500,000-token context window, and multimodal text and image input. Reasoning runs server-side at all times and is controlled via configurable reasoning effort levels, with no visible chain-of-thought tags in the response text. Function calling, structured outputs, and prompt caching round out the profile for coding and agentic workflows.
Gemma 4 31B Instruct is Google’s largest dense model, with 30.7 billion parameters across 60 layers and hybrid attention combining local sliding-window with global attention. Unlike the Gemma 4 MoE variant, this model activates all parameters per token. Multimodality for text and images, 256,000 tokens of context, native function calling, and a configurable thinking mode round out the profile. True Apache 2.0 license.
DeepSeek V3.2 is designed as a Frontier model for language, code, and reasoning, using the same MoE architecture as its predecessor with 671 billion total and 37 billion active parameters. The model operates with a 128,000-token context window, is available as an Open Weights variant for local deployment, and is accessible via cloud API at low prices. The Chinese jurisdiction makes an assessment of cloud deployment necessary.
What most Edge models can’t do: process text, image, audio, and video in a single weight package without a separate multimodal projector file. Gemma 4 E4B uses per-layer embeddings for 4.5 billion effective parameters and runs on Edge hardware with around 5 gigabytes of VRAM. Configurable thinking modes and a 128,000-token context under the Apache 2.0 license round out the profile.
This community fine-tune variant of Qwen 3.6 35B-A3B removes the safety filters and delivers unfiltered responses without Refusals. Of the 35 billion total parameters in the MoE architecture, only 3 billion are active per token; the context window spans 262,000 tokens. Operable locally at near-full quality with Q8 quantization under the Apache 2.0 license, with multimodal processing for text, image, and video.
Kimi K2 is Moonshot AI’s flagship model with an Open Weights license, focused on agentic tasks, coding, and multi-step reasoning. The MoE architecture activates 32 billion of the total one trillion parameters per token; the context window spans 128,000 tokens. Available as a text-only variant locally or via cloud providers — cloud usage under Chinese jurisdiction requires a separate evaluation.
Gemma 4 12B Instruct as a Q8 quantization by the Unsloth community, the highest-precision variant among the 12B builds. With twelve billion parameters and a 128,000-token context window, the model delivers near-FP16 quality, is designed for local operation without cloud connectivity, and is fully commercially usable under the Apache 2.0 license.
Mistral Small 4 is Mistral AI’s compact Open Weights model for general and agentic tasks. The MoE architecture activates only 6.5 billion of the total 119 billion parameters per token, the context window spans 256,000 tokens, and the model processes text and image inputs. Available under the Apache 2.0 license for local use or via the Mistral API, from a European provider environment.
Xiaomi MiMo V2.5 is a natively omnimodal MoE model with 310 billion total and 15 billion active parameters. The model processes text, image, video, and audio within a single architecture; the context window supports up to one million tokens. Fully commercially usable under the MIT license, from a Chinese manufacturer jurisdiction with a corresponding assessment for cloud deployment.
Hermes 4 70B is an open instruct and reasoning model by Nous Research from the Hermes 4 family with 70 billion parameters. The model combines optional thinking with advanced tool use and structured outputs, trained for high steerability and reduced Refusal rates. Available as an Open Weights model under a Modified MIT license for local or server-side deployment.
NVIDIA Llama 3.3 Nemotron Super 49B v1.5 is a pruning- and distillation-optimized variant of Meta’s Llama 3.3 70B with 49 billion parameters. The model delivers strong reasoning performance at reduced resource requirements, a context window of 131,000 tokens, and an optional thinking mode controlled via system prompt. Available as an Open Weights variant under the NVIDIA Open Model License, locally or through cloud providers.
Devstral 2 is Mistral AI’s code agent for demanding software engineering tasks. With 123 billion parameters, the model operates on codebase exploration, multi-file changes, debugging, and legacy modernization, supports a context window of 256,000 tokens, and processes both text and image inputs. Available as an Open Weights model under a modified MIT license, from a European provider environment with GDPR compliance.
GPT-5.4 Mini is the compact GPT-5.4 variant for fast and cost-efficient everyday tasks. With a context window of 272,000 tokens and multimodal input for text and image, the model targets applications requiring low latency with solid output quality. Available exclusively via the OpenAI API.
Grok 4 (Non-Reasoning) is the standard variant of xAI’s Grok 4 generation without active chain-of-thought, designed for fast, direct responses. The model operates with a context window of one million tokens, is available exclusively via the xAI API, and targets conversational and text tasks without explicit reasoning steps.
Grok 4.20 (Reasoning) is xAI’s Frontier model with a dedicated reasoning mode and proprietary cloud-only license from the USA. Released on March 9, 2026, the model delivers reasoning steps via provider metadata rather than content tags and is part of a family that includes non-reasoning and multi-agent variants.
Qwen 3.5 4B is a compact Open Weights model from Alibaba featuring a Thinking-Optional architecture and Unsloth Dynamic quantization at Q6. The Q6 tier offers higher fidelity than the Q4 variant at a moderate VRAM overhead, with a context window of 128,000 tokens. Deployable locally on resource-constrained hardware under the Apache 2.0 license.
o4-mini is OpenAI’s compact reasoning model with native vision input for images, diagrams, and screenshots. The model processes text and image, operates with a context window of 200,000 tokens, and offers three adjustable reasoning levels for balancing response depth and latency. Full tool use including parallel tool calling for lightweight agentic workflows.
Qwen 3 32B is Alibaba’s open weights model for general tasks, reasoning, and coding. With 32 billion parameters and an optional thinking mode, the model operates with a context window of 128,000 tokens and offers a balanced trade-off between performance and efficiency. Available locally or via cloud providers under the Apache 2.0 license.
Laguna S 2.1 has been Poolside’s third consecutive open coding model since July 2026, a MoE with 117.6 billion total and only 8.5 billion active parameters per token. The Linux Foundation’s OpenMDW-1.1 license permits free commercial use and modification. The model supports 1,024,000 tokens of context, interleaved thinking with tool calls within the chain of thought, and native tool calling for long-horizon agentic workflows.
Gemma 3 12B Instruct as a Q4 quantization, optimized for local inference on resource-constrained hardware. The model processes text and image inputs with a context window of 128,000 tokens, is designed for direct task execution, and operates without an external cloud connection. Licensed under the Google Gemma Terms of Use, which permit commercial use.
GPT-5.4 Nano is OpenAI’s most affordable GPT-5.4 variant for high-volume standard tasks such as classification, extraction, and ranking. With a context window of 272,000 tokens and up to 128,000 tokens of output, the model is well-suited for batch processing and sub-agent routing. Available exclusively via the OpenAI API at low cost.
Qwen 3.5 9B is an Open Weights model by Alibaba for general language and reasoning tasks, featuring a thinking-optional architecture. The Unsloth Dynamic Q6 quantization delivers near-lossless quality at significantly reduced VRAM requirements; the context window spans 128,000 tokens. Runs under the Apache 2.0 license on local hardware without cloud dependency.
NVIDIA Nemotron 3 Nano 30B A3B is an efficient hybrid model from the Nemotron-3 series, combining Mamba-2 with Transformer layers. With 31.6 billion total parameters, the model activates only 3.2 billion per token; the context window supports up to one million tokens. Optional thinking mode with configurable budget, native tool calls, and agentic capabilities out of the box. Available as an Open Weights model under the NVIDIA Open Model License.
Hermes 4 14B Abliterated is a locally deployable Open Weights variant by NousResearch based on Qwen 3 14B, with safety mechanisms deliberately removed. With 14 billion parameters and a 128,000-token context window, the model targets unfiltered response behavior, creative writing workflows, and scenarios requiring maximum openness. Fully commercially usable under the Apache 2.0 license.
o3-mini is OpenAI’s compact reasoning model with internal chain-of-thought, specialized in math, coding, and STEM tasks. The model operates with a context window of 200,000 tokens and offers three adjustable reasoning levels for balancing response depth, latency, and cost. Available exclusively via the OpenAI API.
Hermes 4.3 36B (NVFP4) is the compressed variant of NousResearch’s Open Weights model based on ByteDance Seed-OSS-36B, fully decentralized post-trained via the Psyche network. Fireworks compressed the BF16 checkpoint to NVFP4, significantly reducing the memory footprint of the weights. 36 billion dense parameters, 512,000 tokens of native context, optional thinking mode, and an Apache 2.0 license for free commercial use.
GPT-4o is OpenAI’s multimodal all-around model with native support for text, image, and audio inputs. It operates with a context window of 128,000 tokens, is available exclusively via the OpenAI API, and targets a broad range of productive applications — from analysis and coding to natural conversation.
Hermes 4 405B is a high-performance instruct and reasoning model from Nous Research with 405 billion parameters, designed for complex reasoning tasks and agentic workflows. The model supports optional thinking, precise tool calls, and structured outputs. Trained for high steerability and reduced Refusal rates. Available as an Open Weights model under the Meta Llama Community License.
Qwen 3 14B is Alibaba’s open-weights model for general language tasks and reasoning with an optional thinking mode. The Q6 quantization is designed for efficient local operation without a cloud connection; the context window covers 128,000 tokens. Fully commercially usable under the Apache 2.0 license.
OpenAI o1 is OpenAI’s first dedicated reasoning model, trained on internal chain-of-thought and designed for complex math, logic, and science tasks. The model operates with a context window of 200,000 tokens, is available exclusively via the OpenAI API, and is particularly well-suited for use cases where thorough multi-step reasoning matters more than low latency.
Hermes 4 14B as a Q4 quantization of the NousResearch distribution based on Qwen-3, optimized for local assistance and agentic tasks. With 14 billion parameters and a 128,000-token context window, the model runs on resource-constrained hardware and supports hybrid reasoning modes. Fully commercially usable under the Apache 2.0 license.
Grok 4.3 is xAI’s current flagship model with native real-time access to the X platform and a context window of one million tokens. The MoE architecture enables efficient inference on agentic routine tasks, but falls short of specialized models in complex software engineering. Available exclusively via the xAI API at $1.25 per million input tokens and $2.50 per million output tokens.
Hermes 4.3 36B (NVFP4) is the compressed variant of NousResearch’s Open Weights model based on ByteDance Seed-OSS-36B, fully decentralized post-trained via the Psyche network. Fireworks compressed the BF16 checkpoint to NVFP4, significantly reducing the memory footprint of the weights. 36 billion dense parameters, 512,000 tokens of native context, optional thinking mode, and an Apache 2.0 license for free commercial use.
Trained for code tasks, not for general use: Codestral 25.08 is Mistral AI’s specialized developer model from August 2025, optimized for fill-in-the-middle and a broad range of programming languages. With 22 billion parameters, it supports 128,000 tokens of context, runs either locally or via the Mistral API, but is subject to a restrictive license with limited commercial use.
GPT-4o Mini is OpenAI’s compact entry-level model in the GPT-4o family, designed for low cost and fast response times. With a context window of 128,000 tokens, the model processes text and image inputs, is available exclusively via the OpenAI API, and is suited for everyday tasks such as classification, simple text generation, and cost-efficient automation.
Llama 3.3 70B Versatile is Meta’s all-round model in the 70-billion class, with balanced strengths across a broad range of tasks. With a 128,000-token context window and Open Weights under the Llama 3.3 Community License, the model is available either locally for maximum data sovereignty or through cloud providers.
Cohere’s first open MoE model combines frontier-level performance with self-hosting: of 218 billion total parameters, only 25 billion are active per token, the context window spans 128,000 tokens, and the model processes both text and image inputs. Released under the Apache 2.0 license out of Canada, with no vendor lock-in and no restricted weights.
Four billion parameters, Apache 2.0 license, and an optional thinking mode: Qwen 3 4B is designed for mobile applications and edge setups where every watt counts. Locally deployable under Q6 quantization, with a 128,000-token context window — generously sized for a Nano-class model. The manufacturer’s jurisdiction of China occasionally results in censored or evasive responses on politically sensitive topics.
Hermes 3 8B is an uncensored fine-tune by NousResearch based on Meta’s Llama 3.1 8B. With eight billion parameters and a 128,000-token context window, the model targets instruction following, tool use, and creative or ambiguous requests. Deployable locally under the Llama 3.1 Community License, the reduced refusal rate is a defining characteristic of this distribution.
GPT-OSS 20B is OpenAI’s compact Open Weights model from the OSS series, featuring 21 billion parameters in a dense architecture, hosted via Groq for cloud use and simultaneously available as a download for local deployment. The model supports configurable reasoning with three intensity levels and an Apache 2.0 license for full commercial usability.
Qwen 2.5 Coder 7B is a compact Open Weights coding model by Alibaba, optimized for code generation, debugging, and repair. Q6 quantization enables local operation on resource-efficient hardware with minimal quality loss; the context window spans 128,000 tokens for complex codebases. Fully commercially usable under the Apache 2.0 license.
Llama 4 Scout is Meta’s multimodal fourth-generation Llama model, combining general language processing with image understanding in an efficient MoE architecture. Of 109 billion total parameters, only 17 billion are active per token; the context window spans 128,000 tokens. Available under the Llama 4 Community License, which contains restrictions for EU-based users regarding self-hosting and deployment.
