Overview of OnlySq models
OnlySq offers a wide range of models for various user needs. All models supports "system" role for more customisation. You can get json of models by using:
GET https://api.onlysq.ru/ai/models
ChatGPT models
ChatGPT is a free and easy-to-use app that can help you with writing, learning, brainstorming, and more.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
gpt-4o-mini | A small model with superior textual intelligence and multimodal reasoning | Keys | Text | API2.0 OpenAI SDK |
gpt-4o | GPT‑4o (“o” for “omni”) is a step towards much more natural human-computer interaction | Provider | Text | API2.0 OpenAI SDK |
gpt-4 | GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses | Provider | Text | API2.0 OpenAI SDK |
o3-mini | OpenAI o3‑mini is first small reasoning model that supports highly requested developer features including function calling(opens in a new window), Structured Outputs(opens in a new window), and developer messages(opens in a new window), making it production-ready out of the gate | Provider | Text | API2.0 OpenAI SDK |
gpt-3.5-turbo | GPT-3.5-turbo is a fast, cost-effective version of GPT-3.5, optimized for dialogue and general tasks with broad capabilities but less powerful than GPT-4. | Provider | Text | API2.0 OpenAI SDK |
DeepSeek models
Deepseek is a series of large language models developed by the company Deepseek, known for their strong performance in code generation, reasoning, and natural language tasks.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
deepseek-r1 | Deepseek-R1 is a specialized AI model by Deepseek, optimized for reasoning and code generation tasks, with strong performance in logical problem-solving and programming. It’s designed to handle complex, multi-step tasks efficiently. | Provider | Text | API2.0 OpenAI SDK |
deepseek-v3 | Deepseek-V3 is an advanced AI model by Deepseek, excelling in reasoning, code generation, and multi-step problem-solving. It builds on prior versions with enhanced capabilities for complex tasks and improved efficiency. | Provider | Text | API2.0 OpenAI SDK |
Llama models
Llama is a series of open-source large language models developed by Meta. Known for their versatility and multilingual support, Llama models excel in natural language understanding, code generation, and reasoning tasks.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
llama-3.3 | Llama-3.3 is an advanced iteration of Meta's Llama series, featuring enhanced reasoning, multilingual support, and improved efficiency for complex tasks like coding and natural language understanding. | Provider | Text | API2.0 OpenAI SDK |
llama-3.1 | Llama-3.1 is an updated version of Meta's Llama series, offering improved performance in natural language understanding, code generation, and multi-language support, with a focus on scalability and efficiency for various tasks. | Provider | Text | API2.0 OpenAI SDK |
Claude models
Claude is a series of large language models developed by Anthropic, known for their strong natural language understanding, reasoning, and code generation capabilities.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
claude-3.5-sonnet | Claude-3.5-Sonnet is a powerful AI model by Anthropic, excelling in natural language tasks, reasoning, and code generation, with a focus on safety, accuracy, and handling complex, multi-step problems. | Provider | Text | API2.0 OpenAI SDK |
claude-3-haiku | Claude-3-Haiku is a compact, fast, and cost-efficient AI model by Anthropic, designed for simple tasks with a focus on speed and affordability while maintaining strong language understanding and reasoning capabilities. | Provider | Text | API2.0 OpenAI SDK |
Le Chat models
Mistral is a series of open-weight large language models developed by Mistral AI, known for their efficiency, scalability, and strong performance in natural language tasks, reasoning, and code generation.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
mistral-small-3.1 | Mistral-Small-3.1 is a compact, efficient variant of the Mistral series, designed for cost-effective and fast performance on simpler tasks while maintaining strong language understanding and reasoning capabilities. | Keys | Text | API2.0 OpenAI SDK |
Qwen models
Qwen is a series of large language models developed by Alibaba Cloud, designed for versatility and high performance across a wide range of tasks.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
qwen3-235b-a22b | A large-scale language model with 235 billion parameters, designed for complex reasoning, multi-step tasks, and high-accuracy applications. | Keys | Text | API2.0 OpenAI SDK |
qwen3-30b-a3b | A powerful LLM with 30 billion parameters, optimized for advanced natural language understanding, code generation, and multi-modal tasks. | Keys | Text | API2.0 OpenAI SDK |
qwen3-32b | A 32-billion-parameter version of Qwen3, offering strong performance in text generation, logical reasoning, and coding tasks. | Keys | Text | API2.0 OpenAI SDK |
qwen-max-latest | The most capable and up-to-date version of Qwen, ideal for handling complex, multi-step tasks with high accuracy. | Keys | Text | API2.0 OpenAI SDK |
qwen-plus-2025-01-25 | A balanced model released on January 25, 2025, offering strong performance across a wide range of tasks at a moderate cost. | Keys | Text | API2.0 OpenAI SDK |
qwq-32b | A research prototype focusing on advanced reasoning capabilities, particularly excelling in math and code generation. | Keys | Text | API2.0 OpenAI SDK |
qwen-turbo-2025-02-11 | A lightweight and fast model released on February 11, 2025, optimized for simple, cost-effective tasks requiring quick responses. | Keys | Text | API2.0 OpenAI SDK |
qwen2.5-omni-7b | A 7-billion-parameter multi-modal model capable of processing both text and images for integrated understanding and generation. | Keys | Text | API2.0 OpenAI SDK |
qvq-72b-preview-0310 | A preview version of a 72-billion-parameter model, featuring enhanced reasoning and generation capabilities for advanced tasks. | Keys | Text | API2.0 OpenAI SDK |
qwen2.5-vl-32b-instruct | A vision-language model with 32 billion parameters, fine-tuned to understand and generate responses from visual inputs like images. | Keys | Text | API2.0 OpenAI SDK |
qwen2.5-14b-instruct-1m | A 14-billion-parameter model trained to handle long-context tasks, supporting input lengths up to 1 million tokens. | Keys | Text | API2.0 OpenAI SDK |
qwen2.5-coder-32b-instruct | A specialized 32B model fine-tuned for code generation and programming-related tasks, supporting multiple programming languages. | Keys | Text | API2.0 OpenAI SDK |
qwen2.5-72b-instruct | A 72-billion-parameter instruction-following model with strong capabilities in natural language understanding and generation. | Keys | Text | API2.0 OpenAI SDK |
qwen | Qwen is a large language model developed by Alibaba Cloud, excelling in natural language understanding, code generation, and multi-language support, with a focus on versatility and handling complex tasks efficiently. | Provider | Text | API2.0 OpenAI SDK |
Google models
Gemini is a series of large language models developed by Google, known for their advanced reasoning, natural language understanding, and multimodal capabilities.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
gemini-2.5-flash | A fast and efficient version of Google's Gemini model, optimized for real-time, low-latency tasks while maintaining strong multi-modal understanding. | Keys | Text | API2.0 OpenAI SDK |
gemini-2.5-flash-lite-preview-06-17 | A lightweight preview release of the Gemini Flash model from June 17, designed for cost-effective, high-speed inference on less complex tasks. | Keys | Text | API2.0 OpenAI SDK |
gemini-2.0-flash | A fast, efficient model optimized for quick responses and cost-effective tasks while maintaining strong performance. | Keys | Text | API2.0 OpenAI SDK |
gemini-2.0-flash-lite | A lighter version of gemini-2.0-flash, prioritizing speed and lower resource usage for simpler tasks. | Keys | Text | API2.0 OpenAI SDK |
gemini-1.5-flash | A fast and capable model from the 1.5 series, designed for general-purpose tasks with a balance of speed and accuracy. | Keys | Text | API2.0 OpenAI SDK |
gemini-1.5-flash-8b | A smaller, 8-billion-parameter variant of gemini-1.5-flash, offering faster inference for less complex tasks. | Keys | Text | API2.0 OpenAI SDK |
gemini-1.5-pro | A high-performance model in the 1.5 series, optimized for complex, multi-step tasks requiring advanced reasoning and precision. | Keys | Text | API2.0 OpenAI SDK |
gemma-3-4b-it | gemma-3-4b-it is a 4-billion-parameter model from the Gemma series, optimized for instructional tasks (IT). It provides strong performance in natural language understanding and task-specific applications, balancing efficiency and capability. | Keys | Text | API2.0 OpenAI SDK |
Cohere models
Command is a specialized AI model developed by Cohere, designed for understanding and executing commands or instructions effectively. It focuses on task-specific applications, such as interpreting user inputs, automating workflows, and performing actions based on natural language commands.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
command-a-03-2025 | A future iteration of the Command model, likely optimized for advanced text generation and conversational tasks. | Keys | Text | API2.0 OpenAI SDK OpenAI SDK |
command-r7b-12-2024 | A specialized version of the Command-R series, potentially focusing on reasoning or command execution with specific improvements. | Keys | Text | API2.0 OpenAI SDK |
command-r-plus-04-2024 | An enhanced variant of Command-R+, designed for superior reasoning and task-specific applications. | Keys | Text | API2.0 OpenAI SDK |
command-r-plus | An advanced version of the Command-R model, emphasizing improved reasoning and execution capabilities. | Keys | Text | API2.0 OpenAI SDK |
command-r-08-2024 | A version of the Command-R model released in August 2024, optimized for interpreting and executing commands. | Keys | Text | API2.0 OpenAI SDK |
command-r-03-2024 | A March 2024 release of the Command-R model, focused on command-driven tasks and automation. | Keys | Text | API2.0 OpenAI SDK |
command-r | The standard Command-R model, tailored for understanding and executing instructions effectively. | Keys | Text | API2.0 OpenAI SDK |
command | The foundational Command model, designed for general-purpose text generation, conversations, and question-answering. | Keys | Text | API2.0 OpenAI SDK |
command-nightly | A nightly build of the Command model, offering the latest experimental features and updates. | Keys | Text | API2.0 OpenAI SDK |
command-light | A lightweight version of the Command model, optimized for speed and efficiency in resource-constrained environments. | Keys | Text | API2.0 OpenAI SDK |
command-light-nightly | A nightly build of the lightweight Command model, providing rapid updates and experimental features for lightweight use cases. | Keys | Text | API2.0 OpenAI SDK |
c4ai-aya-expanse-32b | A large-scale model (32 billion parameters) developed by C4AI, likely designed for expansive reasoning, complex tasks, and high-performance applications. | Keys | Text | API2.0 OpenAI SDK |
Image models
Image Models are AI models designed for tasks involving images, such as recognition, segmentation, generation, enhancement, and style transfer.
| Model name | Description | Type | Modality | Endpoints |
|---|---|---|---|---|
Kandinsky | Kandinsky is a series of multimodal AI models developed by Sber, designed for generating and processing images based on textual descriptions. | Keys | Images | ImaGen |