AdaL

    AdaL Portal

    One plan. A full model team.

    Choose the planner that runs the agent loop. AdaL can then bring in separate models for images, video, voice, vision, reading, and other specialist work when the task needs them.

    Explore the model catalog

    Planner models

    Newest planner models

    Planner models reason, call tools, inspect results, and decide the next step.

    Anthropic

    Claude Fable 5.1

    Claude Fable 5.1 for demanding reasoning and long-horizon agentic work in AdaL.

    OpenAI

    GPT-6 Astra

    GPT-6 Astra for the hardest end-to-end reasoning and coding work in AdaL.

    OpenAIDEFAULT

    GPT-6.1 Sol

    GPT-6.1 Sol for complex coding, computer use and agentic work in AdaL, at $2/M input and $10/M output.

    Google

    Gemini 3.8 Flash

    Gemini 3.8 Flash for coding and agentic workflows with vision and 1M context.

    Google

    Gemini 4 Argon

    Gemini 4 Argon, Google's flagship reasoning model, for complex coding and multi-step agent work in AdaL, with 1M context.

    DeepSeek

    DeepSeek V4 Flash Vision

    Experimental vision-capable DeepSeek V4 Flash for visual coding workflows in AdaL.

    Meta

    Muse Wonder

    Muse Wonder for long-horizon agentic coding with effort up to max and 1M context in AdaL.

    Meta

    Muse Spark 1.3

    Muse Spark 1.3 for long-horizon agentic coding with vision and 1M context in AdaL.

    Anthropic

    Claude Sonnet 5.5

    Claude Sonnet 5.5 for coding, investigation and review in AdaL, at $2/M input and $10/M output.

    Anthropic

    Claude Sonnet 5

    Claude Sonnet 5 for coding and agents. Available in AdaL CLI and Desktop.

    OpenAI

    GPT-5.6 Terra

    GPT-5.6 Terra balances intelligence and cost for everyday coding in AdaL.

    Z.ai

    GLM-5.3 Flash

    Z.ai GLM-5.3 Flash — native multimodal, flash-cost coding model with 1M context and vision support in AdaL.

    Alibaba

    Qwen3.8 Flash

    Run Alibaba's Qwen3.8 Flash in AdaL: a multimodal MoE model and early Qwen4-architecture preview with 125B total / 6B activated parameters, 262K native context (1M with YaRN), and $0.16/M input · $0.47/M output. Pick it with /model.

    Portal specialist models

    Image generation

    AdaL can route visual work to dedicated image models without changing the planner running the task.

    GoogleRECOMMENDED

    Nano Banana 2.1

    Fast image generation and multi-turn editing with sharper text, extreme aspect ratios, and output up to 4K.

    Used by tools, not selected as the planner

    GoogleRECOMMENDEDDEFAULT

    Nano Banana 2

    All-around image generation with instruction following, reference images, and output up to 4K.

    Used by tools, not selected as the planner

    GoogleRECOMMENDED

    Nano Banana Pro

    Professional image assets, stronger text rendering, and high-resolution output up to 4K.

    Used by tools, not selected as the planner

    OpenAIRECOMMENDED

    GPT Image 2.5 Flare

    Fast, high-fidelity image generation and editing with reference-image input and text rendering.

    Used by tools, not selected as the planner

    OpenAIRECOMMENDED

    GPT Image 2.5 Sunburst

    Precise image editing and generation for polished final assets, with reference-image input.

    Used by tools, not selected as the planner

    Portal specialist models

    Video generation

    Dedicated video models handle text-to-video, image-to-video, reference-driven shots, native audio, and higher-resolution renders.

    GoogleRECOMMENDED

    Gemini Omni Flash

    Fast text-to-video and image-to-video generation with native audio, 3-10 second clips from 360p drafts to 4K.

    Used by tools, not selected as the planner

    MiniMax

    MiniMax H3

    Text-to-video and image-to-video generation with reference images, native stereo audio, and output up to 2K.

    Used by tools, not selected as the planner

    BytePlus

    Seedance 2.5

    Text-to-video and image-to-video generation with reference images and native audio, up to 30 seconds at 1080p.

    Used by tools, not selected as the planner

    Portal specialist models

    Voice and audio

    Audio models provide narration, responsive text-to-speech and music scores while the planner continues coordinating the broader task.

    ElevenLabsRECOMMENDED

    ElevenLabs Multilingual v2

    High-quality multilingual narration and text-to-speech generation.

    Used by tools, not selected as the planner

    ElevenLabs

    ElevenLabs Turbo v2.5

    Low-latency narration and multilingual text-to-speech generation.

    Used by tools, not selected as the planner

    ElevenLabs

    ElevenLabs Flash v2.5

    Fast text-to-speech generation for responsive voice workflows.

    Used by tools, not selected as the planner

    ElevenLabsRECOMMENDED

    ElevenLabs Music v2

    Music scores with timed sections for scene cuts.

    Used by tools, not selected as the planner

    Google

    Gemini 3.8 Flash TTS

    Text-to-speech with delivery directed in plain words, and two-person dialogue in one call.

    Used by tools, not selected as the planner

    Supporting intelligence

    Small model roles behind the workflow

    The registry also assigns models to supporting jobs. These roles help AdaL preserve context, inspect visual evidence, read documents, and understand audio and video.

    Compaction

    GPT-6 Luna

    Compresses long sessions while preserving the working context.

    Vision

    GPT-6 Luna

    Analyzes screenshots, images, and visual evidence for the active task.

    Reading

    GPT-6 Luna

    Extracts useful structure and evidence from fetched documents.

    Audio and video

    Gemini 3.8 Flash

    Listens to recordings and watches videos to answer questions about them.