Explore the current fjuja catalog, from local in-browser models that run on your GPU to API-driven models available through OpenRouter.
Run entirely in your browser on your own GPU via WebGPU — no API key, usage charge, or prompt data sent to a model provider.
Under 500 MB on disk. Fast, compact in-browser models for basic instructions, short answers, and lightweight experiments.
Qwen2.5-0.5B-Instruct-q4f16_1-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen2.5-Coder-0.5B-Instruct-q4f16_1-MLC
A compact code-focused option for snippets, explanations, and programming experiments. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen3-0.6B-q4f16_1-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen3.5-0.8B-q4f16_1-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
SmolLM2-135M-Instruct-q0f16-MLC
A very compact SmolLM-family model for lightweight chat and small-model experiments. This full-precision build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
SmolLM2-360M-Instruct-q4f16_1-MLC
A very compact SmolLM-family model for lightweight chat and small-model experiments. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Approximately 500 MB to 1 GB on disk. Small in-browser models for chat and everyday tasks on modest hardware.
gemma3-1b-it-q4f16_1-MLC
A lightweight Google Gemma-family model for general-purpose instruction following. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Llama-3.2-1B-Instruct-q4f16_1-MLC
A compact Llama-family model suited to general chat, summarization, and writing. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
phi-1_5-q4f16_1-MLC
A compact Microsoft model designed to deliver strong reasoning for its size. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
SmolLM2-360M-Instruct-q0f16-MLC
A very compact SmolLM-family model for lightweight chat and small-model experiments. This full-precision build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
TinyLlama-1.1B-Chat-v1.0-q4f16_1-MLC
A compact Llama-family model suited to general chat, summarization, and writing. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Approximately 1 GB to 2 GB on disk. More capable local models with a moderate browser download and memory footprint.
gemma-2-2b-it-q4f16_1-MLC
A lightweight Google Gemma-family model for general-purpose instruction following. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen2.5-0.5B-Instruct-q0f16-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This full-precision build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen2.5-Coder-0.5B-Instruct-q0f16-MLC
A compact code-focused option for snippets, explanations, and programming experiments. This full-precision build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen3-0.6B-q0f16-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This full-precision build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen3-1.7B-q4f16_1-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen3.5-0.8B-q0f16-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This full-precision build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
SmolLM2-1.7B-Instruct-q4f16_1-MLC
A very compact SmolLM-family model for lightweight chat and small-model experiments. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Approximately 2 GB to 3 GB on disk. Larger in-browser models for stronger writing, coding, and instruction following.
Llama-3.2-1B-Instruct-q0f16-MLC
A compact Llama-family model suited to general chat, summarization, and writing. This full-precision build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Llama-3.2-3B-Instruct-q4f16_1-MLC
A compact Llama-family model suited to general chat, summarization, and writing. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Phi-3-mini-4k-instruct-q4f16_1-MLC
A compact Microsoft model designed to deliver strong reasoning for its size. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Phi-4-mini-instruct-q4f16_1-MLC
A compact Microsoft model designed to deliver strong reasoning for its size. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen2.5-3B-Instruct-q4f16_1-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Approximately 3 GB to 6 GB on disk. The most capable local options, with significant download, GPU, and memory requirements.
gemma-2-9b-it-q4f16_1-MLC
A lightweight Google Gemma-family model for general-purpose instruction following. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Hermes-2-Pro-Llama-3-8B-q4f16_1-MLC
An instruction-tuned model designed for capable chat and structured responses. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Llama-3.1-8B-Instruct-q4f16_1-MLC
A compact Llama-family model suited to general chat, summarization, and writing. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
NeuralHermes-2.5-Mistral-7B-q4f16_1-MLC
An instruction-tuned model designed for capable chat and structured responses. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen3-8B-q4f16_1-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Qwen3.5-9B-q4f16_1-MLC
A compact Qwen-family model with useful instruction-following and multilingual ability. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
WizardMath-7B-V1.1-q4f16_1-MLC
A math-focused model tuned for numerical reasoning and worked solutions. This 4-bit quantized build runs entirely in your browser via WebLLM, so prompts and responses stay on your device.
Run through OpenRouter using its API, with per-million-token input and output prices shown for every model. New fjuja users are offered a shared sample of Chat Tester and image-generation runs. Video generation always requires your own OpenRouter API key, available here: OpenRouter API keys.
Lowest-cost OpenRouter options for quick drafts, extraction, classification, and high-volume comparisons.
ibm-granite/granite-4.0-h-micro
A IBM Granite model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's mini tier and runs through OpenRouter.
View on OpenRouter →meta-llama/llama-3.2-1b-instruct
A Meta model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's mini tier and runs through OpenRouter.
View on OpenRouter →mistralai/ministral-3b-2512
A Mistral AI model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's mini tier and runs through OpenRouter.
View on OpenRouter →Affordable OpenRouter models that balance speed, quality, and cost for everyday work.
amazon/nova-lite-v1
A Amazon model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →amazon/nova-micro-v1
A Amazon model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3.1-flash-lite
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →google/gemma-4-26b-a4b-it
A Google model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →google/gemma-4-31b-it
A Google model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →gryphe/mythomax-l2-13b
A Gryphe model aimed at complex reasoning and high-capability production work. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →ibm-granite/granite-4.1-8b
A IBM Granite model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →ibm-granite/granite-4.2-8b
A IBM Granite model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →inclusionai/ling-3.0-flash-fin
A Inclusionai model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →meta-llama/llama-3.1-8b-instruct
A Meta model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →meta/muse-glimmer-30b
A Meta model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →meta/muse-spark-1.2-contributor
A Meta model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →microsoft/phi-4-mini-instruct
A Microsoft model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →mistralai/ministral-14b-2512
A Mistral AI model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →mistralai/ministral-8b-2512
A Mistral AI model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →nvidia/nemotron-3-nano-30b-a3b
A Nvidia model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →nvidia/nemotron-3.5-lightning
A Nvidia model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.4-nano
A OpenAI model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-oss-20b
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →poolside/laguna-xs-2.1
A Poolside model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3-32b
A Qwen model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3-coder-30b-a3b-instruct
A Qwen model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3-coder-next
A Qwen model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.6-27b
A Qwen model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.6-35b-a3b
A Qwen model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.8-27b
A Qwen model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →tencent/hunyuan-a13b-instruct
A Tencent model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →z-ai/glm-5.3-flash
A Z.AI model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's lightweight tier and runs through OpenRouter.
View on OpenRouter →Stronger general-purpose OpenRouter models for writing, analysis, coding, and more demanding prompts.
amazon/nova-2-lite-v1
A Amazon model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →anthropic/claude-haiku-4.5
A Anthropic model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →bytedance-seed/seed-2.0-lite
A ByteDance Seed model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →deepseek/deepseek-v4-flash
A Deepseek model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →deepseek/deepseek-v4.1-flash
A Deepseek model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →inception/mercury-2
A Inception model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →meta-llama/llama-4-scout
A Meta model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →mistralai/mistral-small-2603
A Mistral AI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →nvidia/nemotron-3-super-120b-a12b
A Nvidia model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.4-mini
A OpenAI model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.6-luna
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.6-luna-pro
A OpenAI model aimed at complex reasoning and high-capability production work. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-oss-120b
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →poolside/laguna-s-2.1
A Poolside model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3-coder-flash
A Qwen model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.7-plus
A Qwen model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.8-flash
A Qwen model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →stepfun/step-3.7-flash
A Stepfun model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's midrange tier and runs through OpenRouter.
View on OpenRouter →High-capability OpenRouter models for complex reasoning, coding, and polished production work.
ai21/jamba-large-1.7
A AI21 model designed for general-purpose work and long-context document tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →aion-labs/aion-3.0
A Aion Labs model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →amazon/nova-premier-v1
A Amazon model aimed at complex reasoning and high-capability production work. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →amazon/nova-pro-v1
A Amazon model aimed at complex reasoning and high-capability production work. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →anthropic/claude-sonnet-4.6
A Anthropic model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →anthropic/claude-sonnet-5
A Anthropic model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →cohere/command-a
A Cohere model built for enterprise writing, retrieval, and tool-assisted workflows. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3.5-flash-lite
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3.7-flash
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →meta-llama/llama-4-maverick
A Meta model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →mistralai/devstral-2512
A Mistral AI model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →mistralai/mistral-large
A Mistral AI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →mistralai/mistral-medium-3-5
A Mistral AI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →mistralai/mistral-medium-3.1
A Mistral AI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.6-terra
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.6-terra-pro
A OpenAI model aimed at complex reasoning and high-capability production work. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →poolside/laguna-m.1
A Poolside model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.5-397b-a17b
A Qwen model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →tencent/hy3
A Tencent model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →tencent/hy4-preview
A Tencent model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →x-ai/grok-4.6
A xAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →xiaomi/mimo-v2.5-pro
A Xiaomi model aimed at complex reasoning and high-capability production work. It is listed in fjuja's flagship tier and runs through OpenRouter.
View on OpenRouter →The most capable OpenRouter models in the catalog for difficult, high-stakes, and deeply reasoned tasks.
anthropic/claude-fable-5
A Anthropic model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →anthropic/claude-fable-5.1
A Anthropic model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →anthropic/claude-opus-4.8
A Anthropic model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →anthropic/claude-opus-5
A Anthropic model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →anthropic/claude-opus-5-fast
A Anthropic model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →deepseek/deepseek-v4-pro
A Deepseek model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3.1-pro-preview
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3.8-flash
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →minimax/minimax-m3
A Minimax model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →moonshotai/kimi-k2.7-code
A Moonshotai model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →moonshotai/kimi-k3
A Moonshotai model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →nvidia/nemotron-3-ultra-550b-a55b
A Nvidia model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.5
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.5-pro
A OpenAI model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.6-sol
A OpenAI model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.6-sol-pro
A OpenAI model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-6-astra
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-6-astra-pro
A OpenAI model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3-coder-plus
A Qwen model optimized for software development, code generation, and technical problem solving. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.7-max
A Qwen model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen3.8-max
A Qwen model aimed at complex reasoning and high-capability production work. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →z-ai/glm-5.2
A Z.AI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →z-ai/glm-5.3
A Z.AI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's frontier tier and runs through OpenRouter.
View on OpenRouter →Verified OpenRouter image-output models for generation and source-image editing.
black-forest-labs/flux.2-pro
A Black Forest Labs model aimed at complex reasoning and high-capability production work. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →bytedance-seed/seedream-4.5
A ByteDance Seed model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3.1-flash-image
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3.1-flash-lite-image
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →google/gemini-3-pro-image
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →krea/krea-2-large
A Krea model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →krea/krea-2-medium
A Krea model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →microsoft/mai-image-2.6
A Microsoft model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-image-2
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →openai/gpt-5.4-image-2
A OpenAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →qwen/qwen-image-3
A Qwen model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →sourceful/riverflow-v2.5-fast
A Sourceful model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →sourceful/riverflow-v2.5-pro
A Sourceful model aimed at complex reasoning and high-capability production work. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →x-ai/grok-imagine-image-2.0
A xAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's image tier and runs through OpenRouter.
View on OpenRouter →OpenRouter video generation with audio explicitly disabled.
alibaba/wan-3.0
A Alibaba model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →alibaba/wan-3.0-prime
A Alibaba model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →bytedance/seedance-1.5-pro
A Bytedance model aimed at complex reasoning and high-capability production work. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →bytedance/seedance-2.0
A Bytedance model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →bytedance/seedance-2.0-fast
A Bytedance model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →bytedance/seedance-2.0-mini
A Bytedance model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →bytedance/seedance-2.5
A Bytedance model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →google/veo-3.1
A Google model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →google/veo-3.1-fast
A Google model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →google/veo-3.1-lite
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →x-ai/grok-imagine-video
A xAI model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video only tier and runs through OpenRouter.
View on OpenRouter →OpenRouter video generation with native synchronized audio enabled.
bytedance/seedance-1.5-pro
A Bytedance model aimed at complex reasoning and high-capability production work. It is listed in fjuja's video with audio tier and runs through OpenRouter.
View on OpenRouter →bytedance/seedance-2.0
A Bytedance model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video with audio tier and runs through OpenRouter.
View on OpenRouter →bytedance/seedance-2.5
A Bytedance model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video with audio tier and runs through OpenRouter.
View on OpenRouter →google/veo-3.1
A Google model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video with audio tier and runs through OpenRouter.
View on OpenRouter →google/veo-3.1-fast
A Google model suited to general writing, analysis, reasoning, and comparison tasks. It is listed in fjuja's video with audio tier and runs through OpenRouter.
View on OpenRouter →google/veo-3.1-lite
A Google model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's video with audio tier and runs through OpenRouter.
View on OpenRouter →minimax/h3
A Minimax model optimized for fast, cost-efficient everyday requests. It is listed in fjuja's video with audio tier and runs through OpenRouter.
View on OpenRouter →