Enum ProviderKind 

pub enum ProviderKind {
    Deepseek,
    DeepseekAnthropic,
    NvidiaNim,
    Openai,
    Atlascloud,
    WanjieArk,
    Volcengine,
    Openrouter,
    Orcarouter,
    XiaomiMimo,
    Novita,
    Fireworks,
    Siliconflow,
    Arcee,
    SiliconflowCN,
    Moonshot,
    Sglang,
    Vllm,
    Ollama,
    OllamaCloud,
    Huggingface,
    Together,
    Qianfan,
    OpenaiCodex,
    Anthropic,
    Openmodel,
    Zai,
    Stepfun,
    Minimax,
    MinimaxAnthropic,
    Deepinfra,
    Sakana,
    LongCat,
    OpencodeGo,
    OpencodeZen,
    Meta,
    Xai,
    Mistral,
    Telecomjs,
    ModelstudioTokenPlan,
    ModelstudioTokenPlanAnthropic,
    ModelstudioCodingPlan,
    ModelstudioCodingPlanAnthropic,
    Antigravity,
    Google,
    Edenai,
    Custom,
}

Variants

Deepseek

DeepseekAnthropic

NvidiaNim

Openai

Atlascloud

WanjieArk

Volcengine

Openrouter

Orcarouter

XiaomiMimo

Novita

Fireworks

Siliconflow

Arcee

SiliconflowCN

Moonshot

Sglang

Vllm

Ollama

OllamaCloud

Huggingface

Together

Qianfan

OpenaiCodex

Anthropic

Openmodel

Zai

Stepfun

Minimax

MinimaxAnthropic

Deepinfra

Sakana

LongCat

OpencodeGo

OpencodeZen

Meta

Xai

Mistral

Mistral AI — la Plateforme (OpenAI-compatible Chat Completions).

Telecomjs

Jiangsu Telecom TokenHub (OpenAI-compatible).

An AI gateway operated by Jiangsu Telecom that speaks the OpenAI Chat Completions wire protocol and serves a broad model catalog; each API key may access a different subset of models.

ModelstudioTokenPlan

Alibaba Cloud Model Studio — Token Plan (OpenAI-compatible Chat Completions).

Token Plan Personal and Team share the same endpoint. Both the OpenAI and Anthropic dialects are available; select the Anthropic dialect via modelstudio-token-plan-anthropic. Pay-as-you-go workspace-id templating is out of scope for v1; use a custom provider for that plan.

ModelstudioTokenPlanAnthropic

Alibaba Cloud Model Studio — Token Plan Anthropic-compatible endpoint.

ModelstudioCodingPlan

Alibaba Cloud Model Studio — Coding Plan (OpenAI-compatible Chat Completions).

ModelstudioCodingPlanAnthropic

Alibaba Cloud Model Studio — Coding Plan Anthropic-compatible endpoint.

Antigravity

Google Antigravity (agy CLI) — consent-gated read-only credential import only; the cloud-code wire protocol is not implemented and requests fail closed with an actionable message.

Google

Google — Gemini OpenAI-compatible endpoint. Its own backend, not an OpenAI alias: thought signatures on tool calls are captured and replayed per Google’s contract.

Edenai

Eden AI — OpenAI-compatible AI gateway (aggregator).

Serves a broad catalog of upstream models under provider/model namespaced wire ids over the OpenAI Chat Completions protocol.

Custom

User-defined OpenAI-compatible endpoint (#1519).

A single dynamic identity for arbitrary [providers.<name>] kind="openai-compatible" entries. It speaks the OpenAI Chat Completions wire protocol and carries no built-in base URL/model — the concrete endpoint and model arrive via config (base_url / model) and the route’s base_url_override, never from this static descriptor.

Implementations

impl ProviderKind

pub const ALL: [Self; 42]

Catalog / picker surface: one identity per vendor.

Dual-wire dialect kinds (*Anthropic) and Model Studio plan variants stay on the enum for serde and provider_for_kind, but they are not first-class catalog rows. Plan is mode / base_url; dialect is wire = openai|anthropic on the primary provider config.

pub fn all() -> &'static [Self]

pub fn names_hint() -> String

pub fn as_str(self) -> &'static str

pub fn parse(value: &str) -> Option<Self>

pub fn parse_config_identity(value: &str) -> Option<Self>

Parse a provider identifier for config-table identity — the kind used to look up credentials, model, and base URL in the user’s [providers.*] tables.

parse is catalog identity: legacy dual-wire spellings (deepseek-anthropic, minimax-anthropic, the Model Studio plan/dialect kinds) are aliases of the vendor primary and collapse onto it so pickers show one row per vendor. That collapse must not decide which config table holds the user’s credentials: TOML serde keeps the legacy kind for provider = "deepseek-anthropic", so env (CODEWHALE_PROVIDER) and config set provider must resolve the same way or the user’s own named table is orphaned with the key present.

An exact canonical-id or provider_config_key match across the full registry (including legacy dialect/plan kinds) therefore wins over alias collapse; everything else falls back to parse. Wire-endpoint selection is unaffected: it keys off the resolved kind’s wire config, not this parse.

pub fn is_siliconflow(self) -> bool

pub fn secret_store_slot(self) -> &'static str

Canonical durable-credential slot in the local secret store.

Most providers own a slot named after their id. Variants authenticated by the SAME account share one slot so a single saved key (or logout) applies to the whole family:

• SiliconflowCN shares siliconflow (historical China-endpoint slot, already the TUI/CLI convention).
• The four Alibaba Cloud Model Studio variants share modelstudio-token-plan: one Model Studio account/key authenticates the Token Plan and Coding Plan endpoints in both wire dialects, so per-variant slots produced three bogus “missing key” rows whenever one variant held the key.

pub fn provider(self) -> &'static dyn Provider

Return the built-in metadata entry for this provider.

This is a metadata foundation only; runtime routing still resolves through crate::ConfigToml::resolve_runtime_options.

Trait Implementations

impl Clone for ProviderKind

fn clone(&self) -> ProviderKind

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for ProviderKind

impl Debug for ProviderKind

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

Formats the value using the given formatter.

impl Default for ProviderKind

fn default() -> ProviderKind

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for ProviderKind

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Eq for ProviderKind

impl PartialEq for ProviderKind

fn eq(&self, other: &ProviderKind) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl Serialize for ProviderKind

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for ProviderKind