Model Library
Browse all available AI models
seedance-2-0
Seedance 2.0 is a multimodal controllable video generation model developed by ByteDance’s Seed Team. Launched in early February 2026, it is now integrated into Doubao, Jimeng AI, and Volcano Engine (Model ID: doubao-seedance-2-0-260128), with an accelerated version—Seedance 2.0 Fast—available for low-latency scenarios.
gpt-image-2
OpenAI's latest image generation and editing model with a reasoning thinking mode achieving 99%+ text rendering accuracy, supporting up to 2K resolution, flexible aspect ratios, and multilingual text, deeply integrated into ChatGPT and API.
kling-v3
Kuaishou's unified multimodal video model with native 4K/60fps output, AI Director multi-shot storyboarding, multilingual native audio, and ultimate character consistency, unifying video understanding, generation, and editing in one workflow.
gemini-3.1-pro-preview
Google's frontier reasoning model, leading on novel logic (ARC-AGI-2 77.1%), graduate-level science (GPQA Diamond 94.3%) and agentic reliability, with a 1M-token context and full multimodal input — built for complex software engineering, long-document analysis and autonomous agent workflows.
gemini-3.1-flash-lite-image
Google's efficiency-tier image model for ultra-low latency and low cost, supporting text-to-image, interleaved generation-editing, and multi-turn local edits with SynthID+C2PA watermarks; ideal for high-volume interactive apps and prototyping.
claude-opus-4-8
Anthropic's flagship LLM for agentic coding, multidisciplinary reasoning, and knowledge work, with sharper judgment and stronger honesty, a 1M-token context window, and fast mode—built for code migration, deep research, and legal/financial analysis.
deepseek-v4-flash
DeepSeek's open-source efficiency-tier flagship — 284B MoE with 13B active parameters, 1M context, dual thinking/non-thinking modes, priced at a fraction of closed frontier models for top price-performance.
gpt-5.5
OpenAI's natively omnimodal, agentic model processing text, image, audio, and video end-to-end, with a 1M-token context window and native computer use, leading coding and knowledge-work benchmarks for autonomous agents and research.
gpt-5.4-mini
OpenAI's most capable mini model yet for coding, computer use, and subagents; supports text and image input, a 400K context window, and extensive tool calling, optimized for low-latency, high-volume production workloads.
claude-opus-4-7
Anthropic's flagship LLM excelling at long-horizon agents and coding, scoring 87.6% on SWE-bench Verified, with 1M context, adaptive thinking, and high-res vision — for complex coding and enterprise workflows.
claude-opus-4-6
Anthropic's frontier flagship model built for deep reasoning, long-horizon agentic coding, and enterprise knowledge work — tops Terminal-Bench 2.0 and HLE with 1M context and adaptive thinking for the hardest tasks.
claude-sonnet-4-6
Anthropic's high-value mid-tier model approaching Opus-level performance on coding, computer use, and agent orchestration, with 1M context and adaptive thinking — built for scaled coding, desktop automation, and enterprise agents.
gemini-3-pro-image
Google's flagship image generation and editing model built on Gemini 3 Pro, featuring up to 4K resolution, precise multilingual text rendering, real-time Google Search grounding, and studio-quality creative controls.
Z.aiglm-5.1
Z.ai's open-source flagship LLM with 754B MoE architecture, purpose-built for agentic coding and long-horizon tasks, capable of 8-hour autonomous execution and topping SWE-Bench Pro.
happyhorse-1.0
Alibaba ATH Innovation Unit's 15B video model — the first with native joint audio-video generation including dialogue, ambient sound, and lip-sync — #1 on the Artificial Analysis Video Arena, built for short-form, ads, and dialogue-driven video.
deepseek-v4-pro
DeepSeek's open-weight flagship 1.6T MoE model with hybrid sparse attention, 1M-token context, competition-grade coding (LiveCodeBench 93.5), and ultra-low pricing — built for code, reasoning, and long-horizon agents.
kimi-k2.6
Moonshot AI's open-weight 1T-parameter MoE multimodal agentic model, excelling in long-horizon coding, 300-sub-agent swarm orchestration, and native tool use — built for autonomous software engineering and agent workflows.
kimi-k3
Moonshot AI's open flagship LLM: a 2.8T-parameter MoE (first open 3T-class model) with Kimi Delta Attention and Attention Residuals, native vision, 1M-token context, and always-on thinking for long-horizon coding, knowledge work, and reasoning.
gemini-3.1-flash-image-preview
Nano Banana (gemini-3.1-flash-image-preview) is a highly multifunctional generative model designed for speed, accuracy, and creative flexibility. It excels at transforming text prompts into stunning visual effects while providing advanced composition and style transfer capabilities.
gemini-3-pro-image-preview
Google's flagship image generation and editing model built on Gemini 3 Pro, featuring up to 4K resolution, precise multilingual text rendering, real-time Google Search grounding, and studio-quality creative controls.
gemini-2.5-flash-image
Nano Banana (`gemini-2.5-flash-image`) is Google's natively multimodal image generation and editing model designed for speed and efficiency, supporting text-to-image, conversational editing, multi-image composition, and character consistency.
seedream-5.0-lite
ByteDance's intelligent image model with Chain of Thought visual reasoning and real-time web search, comprehensively upgraded in understanding, reasoning, and generation, supporting up to 4K output and 14-image reference editing.
seedream-4.0
ByteDance's multimodal image creation engine unifying text-to-image generation and editing in one architecture, supporting up to 4K output, 30+ art style switching, and multi-reference batch generation with 10x faster inference.
seedream-4.5
ByteDance's unified image generation and editing model with precise text rendering, native 4K resolution, and multi-image reference editing, supporting professional typography, material fidelity, and brand visual consistency for commercial design.
MiniMax-Hailuo-2.3
MiniMax's upgraded video model built on Hailuo 02, enhancing complex body motion, facial micro-expressions, and physical realism, with expanded style support including anime, ink wash, and game CG — better quality at the same price.
MiniMax-Hailuo-02
MiniMax's cinematic AI video generation model built on its proprietary NCR architecture with 2.5x efficiency gains, supporting native 1080p output, extreme physics simulation, and precise instruction following, ranked among the top global video models.
wan2.6
Alibaba's multimodal video generation model series supporting role-play (reference-to-video), multi-shot narrative, audio-visual sync, and up to 15-second output, enabling creators to star in AI videos with their own appearance and voice.
gpt-5.6-sol
OpenAI's flagship frontier language model designed for complex reasoning, professional knowledge work, advanced agents, and software engineering, featuring long-context processing, multimodal understanding, and powerful tool capabilities.
gpt-5.6-luna
OpenAI's cost-optimized general-purpose language model designed for high-volume AI workloads, supporting long-context reasoning, multimodal understanding, and agent tooling for enterprise automation, assistants, content generation, and scalable applications.
gpt-5.6-terra
OpenAI's next-generation general-purpose model balancing intelligence and cost, featuring long-context reasoning, multimodal understanding, and full agent tooling for enterprise AI, coding, automation, and professional knowledge work.
gemini-3.5-flash
A frontier-level ultra-fast native five-modal large model developed by Google DeepMind, equipped with dynamic multi-level deep thinking, stable agent orchestration, flagship coding capability and million-token long context, delivering frontier intelligence at half the cost, suitable for enterprise-scale agent clusters, full-stack R&D, bulk multimedia processing, massive document governance and high-concurrency commercial API scenarios.
gpt-5.4-pro
OpenAI's enhanced-reasoning tier built on GPT-5.4's architecture, applying extra compute for higher accuracy on complex, high-stakes tasks, with a million-token context and deep chain-of-thought for professional research and long documents.
gemini-3-flash-preview
A cost-efficient ultra-fast multimodal large model developed by Google DeepMind, equipped with dynamically adjustable deep reasoning, native four-modal parsing, million-token long context and stable batch tool calling, optimized for low-latency high-concurrency scenarios including enterprise agents, high-frequency API services, coding assistance, bulk audio-video processing and consumer chat applications。
gemini-3.1-flash-lite
An ultra-low-cost, ultra-fast native four-modal large model developed by Google DeepMind, equipped with adaptive dynamic reasoning, million-token long context and high-throughput batch tool calling, optimized for massive high-frequency industrial workloads including real-time customer service, bulk text classification & translation, mass document summarization, lightweight agents, receipt data extraction and high-concurrency consumer chat applications.
gemini-3.1-flash-image
Nano Banana (gemini-3.1-flash-image) is a highly multifunctional generative model designed for speed, accuracy, and creative flexibility. It excels at transforming text prompts into stunning visual effects while providing advanced composition and style transfer capabilities.
seedream-5.0-pro
A professional image generation and editing model from ByteDance, built for strong prompt following, text rendering, multi-image reference, and high-fidelity visual detail across commercial design and creative production.
gpt-5.3-codex
OpenAI's agentic coding model merging prior Codex engineering strength with GPT-5.2's reasoning, supporting real-time mid-task steering, setting new highs on terminal and computer-use benchmarks for long-horizon autonomous development work.
gpt-5
OpenAI's unified AI system combining a fast-response model and a deep-reasoning model via real-time routing, delivering state-of-the-art coding, math, multimodal understanding, and health reasoning for chat, development, and agentic tasks.
gpt-5.2
OpenAI's flagship reasoning model for professional work and long-running agents, setting new highs in knowledge work, coding, and long-context understanding, offered in Instant, Thinking, and Pro modes for enterprise automation.
gpt-5.4
OpenAI's unified flagship merging Codex and GPT, the first mainline reasoning model with frontier coding capability, featuring a million-token context and native computer use, leading knowledge-work and coding benchmarks for agentic automation.
gemini-2.5-flash
Google's next-gen lightweight multimodal model featuring ultra-fast inference, native multimodal fusion, and massive context capacity, optimized for high-frequency interactions, real-time responses, and cost-efficient data processing.
gemini-2.5-pro
A flagship native multimodal large model developed by Google DeepMind, equipped with deep thinking reasoning, full-type audio-video-image-text parsing, million-token ultra-long context and enhanced tool calling, suitable for large code repository analysis, scientific mathematical deduction, audio & video content parsing, massive document governance and enterprise agent automation.
minimax-m2.7
MiniMax's self-evolving 230B MoE model (10B active) with native Agent Teams, dynamic tool search, and end-to-end engineering delivery — MLE Bench Lite 66.6% trails only Opus 4.6/GPT-5.4, built for SRE, ML research, and office automation.
sora-2
OpenAI's flagship video and audio generation model built on a diffusion transformer architecture, featuring realistic physics simulation, synchronized audio-visual generation, and multi-shot controllability for cinematic content creation.