Struct ModelCapabilities 

pub struct ModelCapabilities {
    pub supports_tools: bool,
    pub supports_vision: bool,
    pub supports_reasoning: ReasoningCapability,
    pub max_context_tokens: Option<usize>,
    pub max_output_tokens: Option<usize>,
}

Capability flags advertised by a model adapter.

Fields

supports_tools: bool

Model accepts tool/function-calling requests in the chat API.

supports_vision: bool

Model accepts image inputs in messages (vision-capable).

supports_reasoning: ReasoningCapability

Reasoning controls the model exposes — see ReasoningCapability.

max_context_tokens: Option<usize>

Maximum context window in tokens, if known.

max_output_tokens: Option<usize>

The model’s per-response output ceiling in tokens, if known (from /models metadata or a documented per-model table).

Implementations

impl ModelCapabilities

pub fn ollama_default() -> Self

Conservative defaults for an Ollama-served model. We assume tool calling (every modern Ollama-supported model the project targets has it), assume no vision (the safer static default — real vision support is probed from the /api/show capabilities array by OllamaAdapter::vision_supported and refreshed into the runtime snapshot via Msg::ProviderVisionResolved), and treat reasoning as binary on/off (matches the think: bool semantics for everything except gpt-oss).

Trait Implementations

impl Clone for ModelCapabilities

fn clone(&self) -> ModelCapabilities

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ModelCapabilities

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.