Explore Models For Conversational AI Development

GPT Realtime 2.1 is OpenAI's latest realtime reasoning model, with improved alphanumeric recognition, silence and noise handling, and interruption behavior for complex voice-agent workflows.

GPT Realtime 2.1 mini is OpenAI's faster, lower-cost distilled realtime reasoning model, with improved alphanumeric recognition for responsive voice applications.

GPT Realtime 2 is an OpenAI realtime reasoning model with configurable reasoning effort, strong instruction following, and reliable tool use for complex voice-agent workflows.

GPT Realtime 1.5 is OpenAI's flagship audio model for voice agents and customer support, supporting realtime audio and text conversations with tool use.

GPT Realtime mini is OpenAI's cost-efficient realtime model for responsive audio and text conversations over WebRTC, WebSocket, or SIP connections.

GPT-5.6 Sol is OpenAI's newest frontier model for the most complex professional work, leading the GPT-5.6 series with advanced reasoning and the strongest coding performance for high-stakes tasks.

GPT-5.6 Terra is the balanced member of the GPT-5.6 series, pairing near-frontier intelligence with significantly lower cost for everyday professional workloads.

GPT-5.6 Luna is the most cost-efficient member of the GPT-5.6 series, designed for cost-sensitive, high-volume tasks like classification, data extraction, ranking, and sub-agents.

GPT-5.5 is OpenAI's newest frontier model for the most complex professional work, offering stronger coding performance and advanced reasoning for high-stakes tasks.

GPT-5.4 mini is OpenAI's strongest mini model for coding, computer use, and subagents, bringing the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.

GPT-5.4 nano is OpenAI's cheapest GPT-5.4-class model for simple high-volume tasks like classification, data extraction, ranking, and sub-agents.

GPT-5.4 pro is a version of GPT-5.4 that uses more compute to think harder and provide consistently smarter and more precise responses for complex professional tasks.

GPT-5.4 is OpenAI's frontier model for complex professional work, offering the highest capability in the GPT-5 series with a 1M+ context window and advanced reasoning.

GPT-5.2 is the fast, lightweight member of the 5.2 family, optimized for low-latency chat while retaining strong general intelligence.

GPT-5.1 is the latest frontier-grade model in the GPT-5 series, offering stronger general-purpose reasoning, improved instruction adherence, and a more natural conversational style compared to GPT-5.

GPT-5 is the next generation language model with enhanced reasoning capabilities and improved performance across all domains including coding, mathematics, science, and creative tasks.

GPT-5 Mini is the cost-efficient version of GPT-5, offering excellent performance for most tasks while being faster and more affordable than GPT-5.

GPT-5 Nano is the most lightweight and fastest model in the GPT-5 family, optimized for simple tasks requiring quick responses with minimal computational overhead.

o4-mini is the latest small o-series model. It's optimized for fast, effective reasoning with exceptionally efficient performance in coding and visual tasks.

o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels at technical writing and instruction-following.

o3-mini is a cost-efficient reasoning model that's optimized for coding, math, and science, and supports tools and Structured Outputs.

o1 is a powerful reasoning model that supports tools, Structured Outputs, and vision. The model has 200K context and an October 2023 knowledge cutoff.

GPT-4.1 nano is the fastest, most cost-effective GPT 4.1 model.

GPT 4.1 mini provides a balance between intelligence, speed, and cost that makes it an attractive model for many use cases.

GPT 4.1 is OpenAI's flagship model for complex tasks. It is well suited for problem solving across domains.

GPT-4o mini is OpenAI's most cost-efficient small model that's smarter and cheaper than GPT-3.5 Turbo, and has vision capabilities. The model has 128K context and an October 2023 knowledge cutoff.

GPT-4o is faster and cheaper than GPT-4 Turbo with stronger vision capabilities. The model has 128K context and an October 2023 knowledge cutoff.

GPT-4 Turbo is offered at 128K context with an April 2023 knowledge cutoff and basic support for vision.

The GPT-4 model was built with broad general knowledge and domain expertise.

GPT-3.5 Turbo is fast and inexpensive model for simpler tasks.

GPT-3.5 Turbo is fast and inexpensive model for simpler tasks.

A capable and inexpensive, multi-modal model with great performance across all tasks, with a 1 million token context window, and built for the era of Agents.

A small and cost-efficient multi-modal model, built for at-scale usage with a 1 million token context window.

A capable multi-modal model with great performance across all tasks, with a 1 million token context window, and built for the era of Agents.

Muse Spark 1.2 is a coding-optimized model purpose-built for agentic workflows. It improves on code generation, debugging, and codebase understanding, with a 1M-token context window that handles an entire project in one session.

Muse Spark 1.1 is strongest at agentic performance, tool use, and computer use. It excels at long-running tasks with a 1M-token context window, can delegate execution to parallel sub-agents, and is trained to use computer interfaces across desktop, mobile, and browser environments.

Claude Opus 5 is Anthropic's latest Opus model and a step-change improvement over Opus 4.8. It delivers major gains in agentic coding, professional knowledge work, and long-horizon reasoning, with stronger performance per token across effort levels.

Claude Sonnet 5 is Anthropic's best Sonnet-class model for the combination of speed and intelligence, reaching near-Opus quality on coding and agentic work. It excels at iterative development, complex codebase navigation, long-horizon agentic tasks, and polished professional workflows like document drafting and data analysis.

Claude Opus 4.8 is Anthropic's best generally available model for coding, agentic tasks, and enterprise workflows. It improves on Opus 4.7 for complex multi-step coding work, long-horizon agentic tasks, and professional workflows like document drafting, data analysis, and presentations.

Claude Opus 4.7 is Anthropic's most capable generally available model, built for advanced coding, complex agent workflows, vision tasks, and high-stakes professional work that benefits from sustained reasoning and stronger follow-through.

Sonnet 4.6 is Anthropic's most capable Sonnet-class model yet, with frontier performance across coding, agents, and professional work. It excels at iterative development, complex codebase navigation, end-to-end project management with memory, polished document creation, and confident computer use for web QA and workflow automation.

Claude Opus 4.6 is Anthropic's strongest model for coding and long-running professional tasks, optimized for multi-step agent workflows, large codebases, and sustained reasoning over complex engineering work.

Claude Opus 4.5 is Anthropic’s frontier reasoning model optimized for complex software engineering, agentic workflows, and long-horizon computer use. It offers strong multimodal capabilities, competitive performance across real-world coding and reasoning benchmarks, and improved robustness to prompt injection.

Claude 4.5 Sonnet: advanced Sonnet tuned for agents, long coding and sustained reasoning.

Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering near-frontier intelligence at a fraction of the cost and latency of larger Claude models. Matching Claude Sonnet 4's performance across reasoning, coding, and computer-use tasks, Haiku 4.5 brings frontier-level capability to real-time and high-volume applications.

Claude Opus 4 is benchmarked as the world’s best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows.

Claude Sonnet 4 significantly enhances the capabilities of its predecessor, Sonnet 3.7, excelling in both coding and reasoning tasks with improved precision and controllability.

GPT-5.3-Codex is a specialized coding model from OpenAI, optimized for software engineering workflows and long-running autonomous code tasks.

GPT-5.2-Codex is a specialized version of GPT-5.2 optimized for software engineering and coding workflows. It is designed for both interactive development sessions and long, independent execution of complex engineering tasks.

GPT-5.1-Codex-Max is OpenAI's latest agentic coding model, designed for long-running, high-context software development tasks.

GPT-5.1-Codex-Mini is a smaller and faster version of GPT-5.1-Codex.

GPT-5.1-Codex is a specialized version of GPT-5.1 optimized for software engineering and coding workflows. It is designed for both interactive development sessions and long, independent execution of complex engineering tasks.

GPT-5-Codex is a specialized version of GPT-5 optimized for software engineering and coding workflows. It is designed for both interactive development sessions and long, independent execution of complex engineering tasks.

Gemma 4 31B Instruct is Google DeepMind's multimodal 31B class model with text and image input, text output, configurable reasoning, native function calling, and multilingual support across more than 140 languages.

Devstral 2 is Mistral AI's coding-focused model for agentic software engineering workflows.

Top-tier reasoning for high-complexity tasks. The most powerful model of the Mistral AI family.

Cost-efficient reasoning for low-latency workloads.

Top-tier DeepSeek reasoning and coding for the most demanding production workloads. Built for high-context tasks that benefit from deeper deliberation.

Fast, cost-efficient DeepSeek reasoning for latency-sensitive workloads that still need a massive context window and strong tool use.

Premier reasoning offering powered by DeepSeek R1 with Chain of Thought (CoT) and advanced search grounding.

Premier reasoning offering powered by DeepSeek R1 with Chain of Thought (CoT).

Premier search offering with search grounding, supporting advanced queries and follow-ups.

Lightweight offering with search grounding, quicker and cheaper than Sonar Pro.

GLM-5.3-Flash is Z.AI's native multimodal coding model (320B total, 18B active parameters), built for visual coding and professional workflows spanning code, browser, and graphical interfaces at a fast, low-cost tier.

GLM-5.3 delivers comprehensive advancements in complex software engineering and agent capabilities. It uses the same base model as GLM-5.2, with all improvements driven by post-training.

GLM-5.2 delivers powerful coding capabilities, usable 1M-context support, and continued strengths in long-horizon tasks.

GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks. Unlike previous models built around minute-level interactions, GLM-5.1 can work independently and continuously on a single task for more than 8 hours, autonomously planning, executing, and improving itself throughout the process, ultimately delivering complete, engineering-grade results.

GLM-5V-Turbo is Z.AI's vision-enabled turbo model for design-to-code, visual debugging, and agentic GUI workflows, combining multimodal input with GLM-5 generation reasoning at a faster, lower-cost tier.

GLM-5-Turbo is Z.AI's speed-optimized GLM-5 variant for high-volume agentic pipelines, preserving selectable thinking modes and long-context coding capabilities at lower latency and cost than full GLM-5.

GLM-5 is Z.AI's fifth-generation flagship open-source foundation model with 744B parameters (40B active MoE), engineered for complex systems design and long-horizon agent workflows. It delivers production-grade performance on large-scale programming tasks with advanced agentic planning, deep backend reasoning, and iterative self-correction.

GLM-4.7-Flash is Z.AI's fast, cost-efficient model optimized for coding, agentic workflows, and real-world use on lower hardware budgets. It offers strong performance in coding and multistep reasoning tasks while being more affordable than the flagship GLM-4.7.

GLM-4.7 is Z.AI's latest flagship model, featuring upgrades in two key areas: enhanced programming capabilities and more stable multi-step reasoning/execution. It demonstrates significant improvements in executing complex agent tasks while delivering more natural conversational experiences and superior front-end aesthetics.

GLM-4.5-Air is the lightweight variant of Z.AI's latest flagship model family, purpose-built for agent-centric applications. Like GLM-4.5, it adopts the Mixture-of-Experts (MoE) architecture but with a more compact parameter size. GLM-4.5-Air supports hybrid inference modes, offering a "thinking mode" for advanced reasoning and tool use, and a "non-thinking mode" for real-time interaction.

MiniMax M3 combines a 1M-token context window, native multimodal input, and strong coding and agentic performance for long-horizon software engineering, terminal tool use, and web browsing workflows.

MiniMax M2.7 delivers strong real-world software engineering performance across end-to-end project delivery, log analysis, bug troubleshooting, code security, and machine learning workflows.

MiniMax M2.5 is MiniMax's productivity-focused flagship model optimized for coding, office automation, and multi-step agent workflows with strong benchmark performance and high token efficiency.

Qwen 3.8 Flash is Alibaba's fast, cost-efficient multimodal model for coding, agentic workflows, visual understanding, and long-context tasks over a 991K token context window.

Qwen 3.8 Max is Alibaba's latest flagship model, building on Qwen 3.7 Max with stronger programming, reasoning, and long-horizon agentic execution over a 1M token context window.

Qwen 3.7 Max is Alibaba's next-generation flagship model for the agent-centric era, with strong programming, productivity, long-term autonomous execution, and broad agent-level capabilities.

Compared with the previously released Qwen3-Max and Qwen3.6-Plus, this model features enhanced vibe coding abilities, more efficient coding agent execution, and significantly improved front-end development skills. Additionally, its long-tail knowledge retention has been further upgraded.

Qwen 3.6 Plus is Alibaba's Plus-tier Qwen 3.6 model, building on Qwen3.5-Plus with stronger reasoning, instruction following, and agentic capabilities for long-context workflows.

MiMo V2.5 Pro is Xiaomi's Pro-tier MiMo v2.5 Mixture-of-Experts reasoning model for agentic workflows, software engineering, and long-horizon tasks with a 1.05M token context window.

MiMo M2.5 is Xiaomi's native full-modal MiMo v2.5 model, supporting text, image, video, and audio understanding with strong agentic capabilities and a 1.05M token context window.

Kimi K3 is MoonshotAI's flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window.

Kimi K2.7 Code is MoonshotAI's coding-focused Kimi model for software engineering and agentic coding workflows.

Kimi K2.6 demonstrates particularly strong performance in long-horizon coding tasks and produces professional-grade design with code and vision.

Kimi K2.5 is MoonshotAI's flagship multimodal model with 1 trillion parameters (32B activated), 256K context, and strong performance on coding, reasoning, and agentic workflows.

Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work. It stays with complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished application or work artifact.

Grok 4.5 is SpaceXAI's smartest model, with frontier performance across coding, knowledge work, and STEM.

Gemini 3.7 Flash is the next iteration in the Gemini 3 model family, featuring algorithmic improvements to its core reasoning foundation. It supports customizable thinking configurations to control the mix of quality, cost, and latency.

Gemini 3.5 Flash is Google's latest model, highly optimized for coding proficiency and parallel agentic execution loops. It defaults to medium thinking effort for faster and more cost-efficient responses.

Gemini 3.6 Flash delivers higher quality across coding, agentic workflows, and web development with reduced token consumption and fewer model calls compared to previous model iterations.

Gemini 3 Flash Preview is a high speed, high value thinking model designed for agentic workflows, multi turn chat, and coding assistance.

Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases, improving on Gemini 2.5 Flash Lite across audio input, retrieval, translation, extraction, and code while supporting configurable reasoning levels.

Gemini 3.1 Pro Preview is Google’s next-generation frontier model for high-precision multimodal reasoning across text, image, video, audio, and code with a long-context window.

Any custom model created by the user.

Fast and efficient question and answer model with web search grounding.

This model belongs to the GPT-4o mini family of ChatBotKit models. It is designed for question and answer applications. The model has a token limit of 128000 and provides a balance between cost and quality. It is a custom model based on the gpt model architecture.

This model belongs to the GPT-4o family of ChatBotKit models. It is designed for question and answer applications. The model has a token limit of 128000 and provides a balance between cost and quality. It is a custom model based on the gpt model architecture.

This model belongs to the GPT-4 Turbo family of ChatBotKit models. It is designed for question and answer applications. The model has a token limit of 128000 and provides a balance between cost and quality. It is a custom model based on the gpt model architecture.

This model belongs to the GPT-4 family of ChatBotKit models. It is designed for question and answer applications. The model has a token limit of 8 * ONE_K and provides a balance between cost and quality. It is a custom model based on the gpt model architecture.

This model belongs to the GPT 3.5 Turbo family of ChatBotKit models. It is designed for question and answer applications. The model has a token limit of 4000 and provides a balance between cost and quality. It is a custom model based on the gpt model architecture.

GPT Image 2 is OpenAI's latest image generation and editing model. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs with improved fidelity and editing capabilities.

GPT Image 1.5 is an advanced image generation model with improved quality and capabilities. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs with enhanced fidelity.

GPT Image 1 Mini is a state-of-the-art image generation model. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.

GPT Image 1 is a state-of-the-art image generation model. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.

This model is based on the DALL-E 3 architecture. It is a high-quality model that can generate images from text. It is tunable and offers a balance between cost and quality.

This model is based on the DALL-E 2 architecture. It is a high-quality model that can generate images from text. It is tunable and offers a balance between cost and quality.

Gemini 3.1 Flash Lite Image is Google's fast, lower-cost image generation and editing model optimized for high-volume visual workflows. It provides efficient image creation and iterative edits for cost-sensitive production use.

Gemini 3.1 Flash Image (also known as "Nano Banana 2") is Google's latest image generation and editing model delivering Pro-level visual quality at Flash speed. It combines advanced contextual understanding with fast, cost-efficient inference for complex image generation and iterative edits.

Gemini 3 Pro Image (also known as "Nano Banana Pro") is Google's advanced native image generation model for professional and creative workflows, with accurate diagram labeling, web-search-grounded imagery, higher resolution output, and stronger multi-image compositing.

Gemini 2.5 Flash Image (also known as "Nano Banana") is Google's state-of-the-art multimodal image generation model. It excels at text-to-image and image-to-image generation with high quality and fast inference.

Grok Imagine Image 2.0 is xAI's latest image generation model available through Vercel AI Gateway. It generates high-quality images from text prompts, with a higher-resolution 2048x2048 output tier.