🆓 Free

This is a 32-billion parameter model emphasizing reasoning capabilities. The system excels at deep reasoning, complex multi-step logic, and advanced instruction following. Version 3.1 represents impro...

The **Auto Free Collection** is a dynamic system-level collection that automatically routes inference requests to one of the predefined free models available in the system. This collection provides co...

Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to Gemini Flash 1.5, while maintaining quality on par with larger models like Gemini Pro 1.5. It introduces notable e...

Preview of Gemini 2.5 Flash.

Fast Gemini 3 for speed and efficiency.

Lightweight Gemini 3 optimized for mobile devices.

Latest flagship Gemini model with advanced reasoning and multimodal capabilities.

Experimental advanced reasoning engine for complex tasks.

Devstral 2 is a code generation and understanding model for programming tasks. Developed by Mistral AI, this model is optimized for its specific use case category.

Devstral Small 2 is a code generation and understanding model for programming tasks. Developed by Mistral AI, this model is optimized for its specific use case category.

Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window.

Embeddinggemma:300m optimized for generating high-quality embeddings. This model supports multimodal capabilities including vision and image understanding.

Functiongemma:270m is a capable language model from Ollama for general-purpose text generation and analysis tasks. This model supports multimodal capabilities including vision and image understanding....

Gemma3:1b is a capable language model from Ollama for general-purpose text generation and analysis tasks. This model supports multimodal capabilities including vision and image understanding.

Llama 3.1 8B Instruct is Meta's state-of-the-art instruction-tuned language model with 8 billion parameters. It's a compact yet powerful model designed for general-purpose conversational AI, reasoning...

Mistral 7B Instruct is Mistral AI's powerful 7-billion parameter instruction-tuned language model, renowned for exceptional efficiency and speed. Despite having only 7B parameters, it achieves perform...

Qwen2.5 7B Instruct is Alibaba's latest-generation instruction-tuned language model with 7.6 billion parameters, representing a significant upgrade to the Qwen family. Built on 18 trillion tokens of d...

Omni Moderation is a content moderation model for safety and policy compliance checking. Developed by OpenAI, this model is optimized for its specific use case category.

Large embedding model for advanced semantic tasks. This model supports multimodal capabilities including vision and image understanding. It features advanced reasoning capabilities for complex problem...

Efficient text embedding model with high quality representations. This model supports multimodal capabilities including vision and image understanding. It features advanced reasoning capabilities for ...

Legacy embedding model still widely used. This model supports multimodal capabilities including vision and image understanding. It features advanced reasoning capabilities for complex problem-solving ...

Fake Gpt 4 Vision with vision capabilities for processing images and visual content. This model supports multimodal capabilities including vision and image understanding. It features advanced reasonin...

OpenChat is a library of open-source language models fine-tuned with C-RLFT, a strategy inspired by offline reinforcement learning. The model is trained on mixed-quality data without preference labels...

DeepCoder 14B Preview is Agentica's code-focused model optimized for programming tasks and code generation.

QwQ 32B RpR v1 is ArliAI's roleplay-optimized reasoning model combining analytical capabilities with creative character interaction.

Cohere: North Mini Code (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

DeepSeek R1 0528 Qwen3 8B is a distilled version combining R1 reasoning with efficient Qwen3 8B architecture.

Main chat model with JSON output and tool calling capabilities. Optimized for conversational AI, general-purpose tasks, and integration with external tools and APIs.

Main chat model with JSON output and tool calling capabilities. Optimized for conversational AI, general-purpose tasks, and integration with external tools and APIs.

DeepSeek R1 is a reasoning-focused model with extended thinking capabilities for complex multi-step problem solving.

DeepSeek R1 Distill Llama 70B transfers R1 reasoning capabilities to efficient Llama 70B architecture.

Dots Studio: Dots3-Note Preview (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Google: Gemma 4 26B A4B (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Google: Gemma 4 31B (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Google: Lyria 3 Clip Preview is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Google: Lyria 3 Pro Preview is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Free tier version of Alibaba/tongyi Deepresearch 30b A3b:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Allenai/olmo 3 32b Think:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Allenai/olmo 3.1 32b Think:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Arcee Ai/trinity Mini:free. This model supports multimodal capabilities including vision and image understanding.

Your prompt will be processed by a meta-model and routed to one of dozens of models (see below), optimizing for the best possible output.

Free tier version of Cognitivecomputations/dolphin Mistral 24b Venice Edition:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Deepseek/deepseek R1 0528:free. This model supports multimodal capabilities including vision and image understanding.

Google/gemini 2.0 Flash Exp:free with optimized speed for rapid response generation. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Google/gemma 3 12b It:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Google/gemma 3 27b It:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Google/gemma 3 4b It:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Google/gemma 3n E2b It:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Google/gemma 3n E4b It:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Kwaipilot/kat Coder Pro:free. This model supports multimodal capabilities including vision and image understanding. The model is optimized for code generation and programming task...

Free tier version of Meta Llama/llama 3.1 405b Instruct:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Meta Llama/llama 3.2 3b Instruct:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Meta Llama/llama 3.3 70b Instruct:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Mistralai/devstral 2512:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Mistralai/mistral 7b Instruct:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Mistralai/mistral Small 3.1 24b Instruct:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Moonshotai/kimi K2:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Nex Agi/deepseek V3.1 Nex N1:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Nousresearch/hermes 3 Llama 3.1 405b:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Nvidia/nemotron 3 Nano 30b A3b:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Nvidia/nemotron Nano 12b V2 Vl:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Nvidia/nemotron Nano 9b V2:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Openai/gpt Oss 120b:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Openai/gpt Oss 20b:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Qwen/qwen 2.5 Vl 7b Instruct:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Qwen/qwen3 4b:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Qwen/qwen3 Coder:free. This model supports multimodal capabilities including vision and image understanding. The model is optimized for code generation and programming tasks.

Free tier version of Tngtech/deepseek R1t Chimera:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Tngtech/deepseek R1t2 Chimera:free. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Tngtech/tng R1t Chimera:free. This model supports multimodal capabilities including vision and image understanding.

Xiaomi/mimo V2 Flash:free with optimized speed for rapid response generation. This model supports multimodal capabilities including vision and image understanding.

Free tier version of Z Ai/glm 4.5 Air:free. This model supports multimodal capabilities including vision and image understanding.

LiquidAI: LFM2.5-2.6B (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

LongCat Flash Chat is Meituan's efficient model with 131K context optimized for Chinese language tasks.

Llama 3.3 8B Instruct (free) is Meta's efficient instruction-tuned model for general tasks.

A high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total parameters). The mo...

Meta's Llama 4 Scout is a mixture-of-experts (MoE) language model that activates 17 billion parameters from a total of 109 billion parameters. The model supports multimodal inputs (text and images) an...

MAI DS R1 is Microsoft's DeepSeek-based reasoning model with enhanced analytical capabilities.

MiniMax: MiniMax M2.7 (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

MiniMax: MiniMax M3 (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

A 12-billion parameter model featuring a 128k token context window, developed by Mistral in partnership with NVIDIA. The model supports multiple languages including English, French, German, Spanish, I...

Mistral Small is a 22-billion parameter model serving as a convenient mid-point between smaller and larger Mistral options. It emphasizes reasoning capabilities, code generation, and multilingual supp...

Mistral Small is a 22-billion parameter model serving as a convenient mid-point between smaller and larger Mistral options. It emphasizes reasoning capabilities, code generation, and multilingual supp...

NVIDIA: Nemotron 3 Nano Omni (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

NVIDIA: Nemotron 3 Super (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

NVIDIA: Nemotron 3 Ultra (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

NVIDIA: Nemotron 3.5 Content Safety (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

NVIDIA: Nemotron 3.5 Lightning (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Free Models Router is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Ling 3.0 Flash Fin (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:

Qwen2.5 Coder 32B Instruct is a code-focused large language model representing the latest iteration in the Qwen coding series. It replaces the earlier CodeQwen1.5 with significantly enhanced capabilit...

Poolside: Laguna S 2.1 (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Poolside: Laguna XS 2.1 (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Qwen2.5 VL 32B Instruct offers enhanced vision-language capabilities in the 32B configuration.

Qwen3 14B is Alibaba's efficient model with strong general capabilities and 40K context.

Qwen3 30B A3B is an efficient MoE model with 3B active parameters for fast inference.

Thinking Machines: Inkling (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Thinking Machines: Inkling Small (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Z.ai: GLM 5.2 (free) is a capable language model available on LangMart via OpenRouter for general-purpose text generation and analysis tasks.

Stable Diffusion 3.5 Large is the most powerful model in the Stable Diffusion family, featuring superior quality and prompt adherence. It is a Multimodal Diffusion Transformer (MMDiT) text-to-image ge...

Toppy M 7B is a wild 7B parameter model that merges several models using the new `task_arithmetic` merge method from [mergekit](https://github.com/cg123/mergekit). This model combines multiple fine-tu...

tokenizer = AutoTokenizer.from_pretrained("upstage/SOLAR-10.7B-Instruct-v1.0")

MiMo-V2-Flash is an open-source language model developed by Xiaomi featuring a Mixture-of-Experts (MoE) architecture with 309B total parameters and 15B active parameters. It employs hybrid attention m...