AI Engines

Model Compatibility

Owlpen supports a wide range of commercial and open-source foundation models.

Model capability is subject to change based on provider updates.

ManagedOwlpen
BYOKOwlpen
Full All features availableStandard Most functions supportedLimited Basic functions only

List last updated 20 March 2026.

Commercial Models

Tap any row to expand individual model compatibility.

ProviderFamilyOwlpen
AnthropicClaudeFull
OpenAIGPT / o-seriesFull
GoogleGeminiFull
DeepSeekDeepSeekFull
Alibaba CloudQwenFull
xAIGrokFull
CohereCommandStandard
AmazonNovaFull
Mistral AIMistral (API)Full
01.AIYi APIStandard
InflectionPi
AI21 LabsJambaStandard
Reka AIReka

Open-Source Models

Tap any row to expand individual model compatibility.

ProviderFamilyOwlpen
MetaLlama 4Full
MetaLlama 3.xFull
Mistral AIMistral / MixtralFull
MicrosoftPhiStandard
GoogleGemmaStandard
Alibaba CloudQwen (Open)Full
DeepSeekDeepSeek (Open)Full
NVIDIALlama NemotronStandard
MetaCode LlamaStandard
Nous ResearchHermesStandard
CohereAya
TIIFalcon
BigScienceBLOOM
DatabricksDBRXStandard
SnowflakeArctic
01.AIYi (Open)
Allen AIOLMo
xAIGrok (Open)
Stability AIStableLM
InternLMInternLM
Zhipu AIGLM
HuggingFaceSmolLM
BigCodeStarCoder

Compatibility is subject to change. Tier ratings reflect general capability alignment and may vary by task. Smaller models (Tier 3) will experience functionality degradation for complex reasoning, multi-step workflows, and context-heavy tasks. Self-hosted deployments may require additional configuration. All model names, trademarks, and logos are the property of their respective owners. Coaley Peak® Ltd is not affiliated with or endorsed by any third-party model provider.

Model capability is also subject to change based on individual provider service quality. Providers may silently update, retrain, or alter the behaviour of a model under the same version identifier without notice, which can affect output quality, instruction-following, and platform function reliability. Coaley Peak® Ltd does not guarantee that any listed model will continue to perform as documented. This is one reason we support a wide range of model integrations — reducing dependency on any single provider or model version.

← Back to About