Models

MiniMax H3 Image is MiniMax’s image model for text-to-image generation, rendering at 1K, 2K and 4K in aspect ratios from 21:9 to 9:21. It pairs with MiniMax H3 Image Edit for changing and combining images.

Seedream V5.5 is ByteDance Seed’s Seedream 5.5, generating high-quality images at 1K, 1.5K, or 2K with improved realism and text rendering. It pairs with Seedream V5.5 Edit.

Claude Fable 5.2 improves on Claude Fable 5.1 across the board, with the biggest gains in agentic coding, long-running agentic workflows, and knowledge work such as long code refactors, front-end development, and finance and analysis tasks.

DeepSeek V4.1 Flash 1017 Abliterated running in a Trusted Execution Environment (TEE). A text-only reasoning and coding model with refusal behavior removed and a 1M-token context window.

GPT Image 2.6 Flare is the faster tier of OpenAI’s GPT Image 2.6, giving high-fidelity results for campaign and product work at lower latency. It renders at up to 4K and pairs with GPT Image 2.6 Flare Edit.

GPT Image 2.6 Sunburst is the high-fidelity tier of OpenAI’s GPT Image 2.6, made for campaign and product work where detail matters. It renders at up to 4K and pairs with GPT Image 2.6 Sunburst Edit.

Kling 4.0 Flash is a faster version of Kling AI’s Kling 4.0, generating 3- to 20-second clips at 720p from text with native audio.

Krea 3 is the third generation of Krea’s image model, rendering at up to 4K with built-in prompt expansion that fills in detail from short prompts.

DeepSeek V4.1 Flash 1017 Uncensored is an abliterated, uncensored variant of DeepSeek V4.1 Flash 1017, a multimodal Mixture-of-Experts model with 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window.

GPT-6.1 Terra is a fast, efficient model for visual reasoning, precise computer use, and clear, natural writing.

Kling 4.0 is Kling AI’s fourth-generation video model. This text-to-video variant generates 3- to 30-second clips at 720p or 1080p with native audio, and supports multi-shot prompts.

Abliterated Large Flash is Abliteration AI’s unrestricted multimodal model, based on DeepSeek V4.1 Flash and further abliterated. It has 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window.

DeepSeek V4.1 Flash is a multimodal Mixture-of-Experts model with 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window.

DeepSeek V4.1 Flash Uncensored is an abliterated, uncensored variant of DeepSeek V4.1 Flash, a multimodal Mixture-of-Experts model with 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window.

Kimi K3.1 Pro is Moonshot AI’s flagship multimodal reasoning model and the post-trained successor to Kimi K3, with a 1M-token context window.

Kimi K3.1 Pro Highspeed is the faster variant of Kimi K3.1 Pro, Moonshot AI’s flagship multimodal reasoning model.

MiniMax M3.1 Flash is a frontier model for coding, agentic workflows, and complex reasoning, with a 1M-token context window, text, image, and video input, and tunable thinking depth from low to max (max by default).

DeepSeek V4.1 Flash is a multimodal Mixture-of-Experts model with 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window.

Flux 3 Max is Black Forest Labs’ post-trained Flux 3 video model, built for high visual quality at high speed. This text-to-video variant generates 5- to 20-second clips at up to 1080p with synchronized audio.

NVIDIA Nemotron 3.5 Super VL is an open multimodal vision-language model for specialized agents.

Muse Spark Open is an open-weight reasoning model from Meta for coding and agentic workflows.

Gemini 4 Argon is Google’s frontier model for complex workflows across software engineering, enterprise knowledge work, and cybersecurity defense, with a unified 1M token context window and tunable thinking.

Muse Wonder is a frontier reasoning model from Meta for complex personal and professional agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with a 1M-token context window.

Qwen Image 2.1 Turbo is the fast, cost-effective tier of Alibaba’s open-weight Qwen-Image 2.1 series, balancing quality and speed at 1K or 2K. It pairs with Qwen Image 2.1 Turbo Edit.

GPT-6.1 Sol Ultrafast is the same model served in OpenAI’s Ultrafast mode, up to 8x faster. It has a 1.05M token context window (922K input, 128K output), support for text and image inputs, and reasoning effort from low through max.

Claude Haiku 5.5 is built for high-volume, latency-sensitive work such as classification, routing, extraction, and subagent tasks.

MiniMax H3 Max Recast replaces the people in a source video with people from reference photos, preserving the original motion, camera, cuts and soundtrack, at up to 1080p.

Clef (System One) is Cloudflare’s decision model. Like Jev, it answers typed questions about a passage with structured judgments rather than prose: probabilities, choices or rubric scores. It reads up to 32,768 tokens of context.

Nano Banana 2.1 is an update to Google’s Gemini-based Nano Banana 2 image model, generating at up to 4K from prompts of up to 32,768 characters. It pairs with Nano Banana 2.1 Edit.

Abliterated Large V2 is Abliteration AI’s unrestricted reasoning model, based on GLM-5.3 and further abliterated and fine-tuned. Text-only with 1M context, for harder reasoning and evaluation workloads.

Flux 3 is Black Forest Labs’ flagship image model, with native 2K and 4K output, precise layout control and strong text rendering.

Grok Imagine 1.5 Lite is a quicker, more affordable version of xAI’s Grok Imagine 1.5, generating 1- to 15-second clips at up to 1080p from text, with native audio.

Ideogram V4.5 is Ideogram’s typography-focused image model, rendering accurate text for posters, logos and designs at 1K or 2K. It offers low, medium and high quality settings and pairs with Ideogram V4.5 Edit.

Generate expressive text-to-speech audio using ElevenLabs Eleven-v4. High-quality voices with audio tag support and automatic text normalization.

Generate fast text-to-speech audio using ElevenLabs Eleven-v4 Turbo. High-quality low-latency voices with audio tag support and automatic text normalization.

GPT-6.1 Sol delivers near-Astra performance at a lower cost for complex coding, computer use, and professional work.

MiMo-V2.6-Flash is Xiaomi’s efficiency-balanced omnimodal model: a sparse Mixture-of-Experts with 309B total and 15B activated parameters, natively understanding text, image, video, and audio with a 1M-token context window.

Claude Opus 5.5 is a step-change improvement over Opus 5, with the biggest gains in agentic coding, long-running agentic tasks, and knowledge work.

Claude Sonnet 5.5 is a step-change improvement over Sonnet 5. It performs great with well-scoped everyday tasks like building features, fixing bugs, and creating polished documents, slides, and spreadsheets.

Aion 3.5, the successor to Aion 3.0, is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models, with a longer context window.

Aion 3.5 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models.

GPT-6 Luna is OpenAI’s most efficient GPT-6 model for focused, high-volume tasks.

GPT-6 Sol is built for complex coding and agentic workflows.

Qwen Image 2.1 Pro is the full-capacity Qwen-Image 2.1 model from Alibaba, compact and efficient while generating at 1K or 2K. It pairs with Qwen Image 2.1 Pro Edit, which takes up to 10 reference images.

Jev (System One) is TypeSafe AI’s decision model. Instead of writing text, it answers typed questions about a passage with structured judgments: yes-or-no probabilities, choices with confidence, or scores on a rubric.

Grok 4.7, the successor to Grok 4.6, is xAI’s multimodal chat and reasoning model with function calling, structured outputs, adjustable reasoning effort (low/medium/high/xhigh), and a 500K-token context window.

High-quality realistic music generation with optional vocals and configurable duration. Best for polished, production-ready tracks across a wide range of genres.

Latest ElevenLabs music generation with higher-quality realistic tracks, optional vocals, and configurable duration. Best for polished, production-ready songs across a wide range of genres.

MiniMax H3 Max is a post-trained variant of MiniMax’s H3 video model. This text-to-video variant generates 5- to 15-second clips at up to 1080p with native stereo audio.

DeepSeek V4.1 Flash is a multimodal Mixture-of-Experts model with 552B backbone parameters (8B active on input, 16B on output) and a 1M-token context window.

Qwen 3.8 Flash is a multimodal model in the Qwen family, pairing strong reasoning and generation with remarkable speed.

GPT Image 2.5 Flare is the faster tier of OpenAI’s GPT Image 2.5, built for quick, high-quality generation in creator and high-volume work. It renders at up to 4K and pairs with GPT Image 2.5 Flare Edit for scoped edits.

GPT Image 2.5 Sunburst is the premium tier of OpenAI’s GPT Image 2.5, built for precise, production-ready campaign and product imagery. It renders at up to 4K and pairs with GPT Image 2.5 Sunburst Edit for tightly controlled revisions.

Kimi K3 is an ultra-large-scale, open-weight multimodal reasoning model from Moonshot AI.

Wan 3.0 Prime Pro combines the speed of Wan 3.0 Prime with the 4K output of Wan 3.0 Pro, generating 2- to 30-second clips from text, with optional native audio.

Wan 3.0 Pro is the highest-resolution tier of Alibaba’s Wan 3.0, generating 2- to 30-second clips at up to 4K from text, with optional native audio.

GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks.

GLM-5.3 Flash is a reasoning model designed for coding, sustained agentic work, and production workloads. It is suited for long-horizon software engineering, complex reasoning, and workflows that combine text with visual context.

Kimi K2.6 is an open-source, native multimodal agentic model from Moonshot AI with 1T total parameters and 32B active parameters.

Qwen 3.8 27B is a native vision-language dense model with 27B parameters. It improves coding, professional work, research, and long-horizon agentic tasks, with flexible thinking control and image and video understanding.