Skip to content
Demo mode · simulated data

Model Lab · Registry

Model Registry

Open-weight foundation models you can inspect, customize and run on infrastructure you control. Choose a base for your intelligence.

Parameters, context and license reflect public model cards at authoring time; verify before use. Scores and latency are illustrative demo values, not benchmarks.Open weights let you run and adapt a model — they don't transfer ownership of the base model's IP. Your adapters, data and configuration are yours.
  • Llama · Meta

    Llama 3.1 8B Instruct

    Params
    8B
    Context
    128K
    Arch
    Dense
    Llama 3.1 Community License
    reasoning
    71
    groundedness
    78
    EdgeOn-premPrivate cloud
  • Llama · Meta

    Llama 3.3 70B Instruct

    Params
    70B
    Context
    128K
    Arch
    Dense
    Llama 3.3 Community License
    reasoning
    88
    groundedness
    86
    EdgeOn-premPrivate cloud
  • Mistral · Mistral AI

    Mistral 7B Instruct v0.3

    Params
    7.3B
    Context
    32K
    Arch
    Dense
    Apache-2.0
    reasoning
    63
    groundedness
    72
    EdgeOn-premPrivate cloud
  • Mistral · Mistral AI

    Mistral Small 3 (24B)

    Current base
    Params
    24B
    Context
    32K
    Arch
    Dense
    Apache-2.0
    reasoning
    80
    groundedness
    82
    EdgeOn-premPrivate cloud
  • Mistral · Mistral AI

    Mixtral 8x7B Instruct

    Params
    46.7B
    Context
    32K
    Arch
    MoE
    Apache-2.0
    reasoning
    74
    groundedness
    77
    EdgeOn-premPrivate cloud
  • Qwen · Alibaba Cloud (Qwen)

    Qwen2.5 7B Instruct

    Params
    7.6B
    Context
    128K
    Arch
    Dense
    Apache-2.0
    reasoning
    73
    groundedness
    76
    EdgeOn-premPrivate cloud
  • Qwen · Alibaba Cloud (Qwen)

    Qwen2.5 72B Instruct

    Params
    72.7B
    Context
    128K
    Arch
    Dense
    Qwen License Agreement
    reasoning
    87
    groundedness
    84
    EdgeOn-premPrivate cloud
  • Gemma · Google

    Gemma 2 9B IT

    Params
    9B
    Context
    8K
    Arch
    Dense
    Gemma Terms of Use
    reasoning
    72
    groundedness
    75
    EdgeOn-premPrivate cloud
  • Gemma · Google

    Gemma 2 27B IT

    Params
    27B
    Context
    8K
    Arch
    Dense
    Gemma Terms of Use
    reasoning
    79
    groundedness
    79
    EdgeOn-premPrivate cloud
  • Phi · Microsoft

    Phi-3.5 mini Instruct

    Params
    3.8B
    Context
    128K
    Arch
    Dense
    MIT
    reasoning
    64
    groundedness
    69
    EdgeOn-premPrivate cloud
  • OLMo · Ai2

    OLMo 2 13B Instruct

    Params
    13B
    Context
    4K
    Arch
    Dense
    Apache-2.0
    reasoning
    66
    groundedness
    71
    EdgeOn-premPrivate cloud

Search Own Intelligence

Jump to a module, model or dataset