Enum ModelId 

pub enum ModelId {
    Gemini3ProPreview,
    Gemini3FlashPreview,
    GPT5,
    GPT52,
    GPT52Codex,
    GPT5Codex,
    GPT5Mini,
    GPT5Nano,
    GPT51,
    GPT51Codex,
    GPT51CodexMax,
    GPT51Mini,
    CodexMiniLatest,
    OpenAIGptOss20b,
    OpenAIGptOss120b,
    ClaudeOpus46,
    ClaudeOpus45,
    ClaudeOpus41,
    ClaudeSonnet45,
    ClaudeHaiku45,
    ClaudeSonnet4,
    ClaudeOpus4,
    ClaudeSonnet37,
    ClaudeHaiku35,
    DeepSeekChat,
    DeepSeekReasoner,
    HuggingFaceDeepseekV32,
    HuggingFaceOpenAIGptOss20b,
    HuggingFaceOpenAIGptOss120b,
    HuggingFaceDeepseekV32Novita,
    HuggingFaceXiaomiMimoV2FlashNovita,
    HuggingFaceMinimaxM25Novita,
    HuggingFaceGlm5Novita,
    HuggingFaceQwen3CoderNextNovita,
    XaiGrok4,
    XaiGrok4Mini,
    XaiGrok4Code,
    XaiGrok4CodeLatest,
    XaiGrok4Vision,
    ZaiGlm5,
    MoonshotMinimaxM25,
    MoonshotQwen3CoderNext,
    OllamaGptOss20b,
    OllamaGptOss20bCloud,
    OllamaGptOss120bCloud,
    OllamaQwen317b,
    OllamaDeepseekV32Cloud,
    OllamaQwen3Next80bCloud,
    OllamaMistralLarge3675bCloud,
    OllamaQwen3Coder480bCloud,
    OllamaGemini3ProPreviewLatestCloud,
    OllamaDevstral2123bCloud,
    OllamaMinimaxM2Cloud,
    OllamaGlm5Cloud,
    OllamaMinimaxM25Cloud,
    OllamaGemini3FlashPreviewCloud,
    OllamaNemotron3Nano30bCloud,
    MinimaxM25,
    MinimaxM2,
    LmStudioMetaLlama38BInstruct,
    LmStudioMetaLlama318BInstruct,
    LmStudioQwen257BInstruct,
    LmStudioGemma22BIt,
    LmStudioGemma29BIt,
    LmStudioPhi31Mini4kInstruct,
    OpenRouterGrokCodeFast1,
    OpenRouterGrok4Fast,
    OpenRouterGrok41Fast,
    OpenRouterGrok4,
    OpenRouterQwen3Max,
    OpenRouterQwen3235bA22b,
    OpenRouterQwen3235bA22b2507,
    OpenRouterQwen3235bA22bThinking2507,
    OpenRouterQwen332b,
    OpenRouterQwen330bA3b,
    OpenRouterQwen330bA3bInstruct2507,
    OpenRouterQwen330bA3bThinking2507,
    OpenRouterQwen314b,
    OpenRouterQwen38b,
    OpenRouterQwen3Next80bA3bInstruct,
    OpenRouterQwen3Next80bA3bThinking,
    OpenRouterQwen35Plus0215,
    OpenRouterQwen3Coder,
    OpenRouterQwen3CoderPlus,
    OpenRouterQwen3CoderFlash,
    OpenRouterQwen3Coder30bA3bInstruct,
    OpenRouterQwen3CoderNext,
    OpenRouterDeepseekChat,
    OpenRouterDeepSeekV32,
    OpenRouterDeepseekReasoner,
    OpenRouterDeepSeekV32Speciale,
    OpenRouterDeepSeekV32Exp,
    OpenRouterDeepSeekChatV31,
    OpenRouterDeepSeekR1,
    OpenRouterOpenAIGptOss120b,
    OpenRouterOpenAIGptOss120bFree,
    OpenRouterOpenAIGptOss20b,
    OpenRouterOpenAIGpt5,
    OpenRouterOpenAIGpt5Codex,
    OpenRouterOpenAIGpt5Chat,
    OpenRouterAnthropicClaudeSonnet45,
    OpenRouterAnthropicClaudeHaiku45,
    OpenRouterAnthropicClaudeOpus41,
    OpenRouterAmazonNova2LiteV1,
    OpenRouterMistralaiMistralLarge2512,
    OpenRouterNexAgiDeepseekV31NexN1,
    OpenRouterOpenAIGpt51,
    OpenRouterOpenAIGpt51Codex,
    OpenRouterOpenAIGpt51CodexMax,
    OpenRouterOpenAIGpt51CodexMini,
    OpenRouterOpenAIGpt51Chat,
    OpenRouterOpenAIGpt52,
    OpenRouterOpenAIGpt52Chat,
    OpenRouterOpenAIGpt52Codex,
    OpenRouterOpenAIGpt52Pro,
    OpenRouterOpenAIO1Pro,
    OpenRouterStepfunStep35FlashFree,
    OpenRouterZaiGlm5,
    OpenRouterMoonshotaiKimiK20905,
    OpenRouterMoonshotaiKimiK2Thinking,
    OpenRouterMoonshotaiKimiK25,
    OpenRouterMinimaxM25,
}

Centralized enum for all supported model identifiers

Variants

Gemini3ProPreview

Gemini 3 Pro Preview - Preview of next-generation Gemini model

Gemini3FlashPreview

Gemini 3 Flash Preview - Our most intelligent model built for speed, combining frontier intelligence with superior search and grounding

GPT5

GPT-5 - Latest most capable OpenAI model (2025-08-07)

GPT52

GPT-5.2 - Latest flagship general-purpose OpenAI model (2025-12-11)

GPT52Codex

GPT-5.2 Codex - Code-focused GPT-5.2 variant optimized for agentic coding

GPT5Codex

GPT-5 Codex - Code-focused GPT-5 variant using the Responses API

GPT5Mini

GPT-5 Mini - Latest efficient OpenAI model (2025-08-07)

GPT5Nano

GPT-5 Nano - Latest most cost-effective OpenAI model (2025-08-07)

GPT51

GPT-5.1 - Enhanced latest most capable OpenAI model with improved reasoning (2025-11-14)

GPT51Codex

GPT-5.1 Codex - Code-focused GPT-5.1 variant using the Responses API

GPT51CodexMax

GPT-5.1 Codex Max - Maximum context code-focused GPT-5.1 variant

GPT51Mini

GPT-5.1 Mini - Enhanced efficient OpenAI model with improved capabilities (2025-11-14)

CodexMiniLatest

Codex Mini Latest - Latest Codex model for code generation (2025-05-16)

OpenAIGptOss20b

GPT-OSS 20B - OpenAI’s open-source 20B parameter model using harmony

OpenAIGptOss120b

GPT-OSS 120B - OpenAI’s open-source 120B parameter model using harmony

ClaudeOpus46

Claude Opus 4.6 - Next-gen flagship Anthropic model with extended thinking

ClaudeOpus45

Claude Opus 4.5 - Latest flagship Anthropic model with exceptional reasoning (2025-11-01)

ClaudeOpus41

Claude Opus 4.1 - Previous most capable Anthropic model (2025-08-05)

ClaudeSonnet45

Claude Sonnet 4.5 - Latest balanced Anthropic model (2025-10-15)

ClaudeHaiku45

Claude Haiku 4.5 - Latest efficient Anthropic model (2025-10-15)

ClaudeSonnet4

Claude Sonnet 4 - Previous balanced Anthropic model (2025-05-14)

ClaudeOpus4

Claude Opus 4 - Previous flagship model (2025-05-14)

ClaudeSonnet37

Claude Sonnet 3.7 - Latest Claude 3 Sonnet (2025-02-19)

ClaudeHaiku35

Claude Haiku 3.5 - Latest Claude 3 Haiku (2024-10-22)

DeepSeekChat

DeepSeek V3.2 Chat - Fast non-thinking mode

DeepSeekReasoner

DeepSeek V3.2 Reasoner - Thinking mode with structured reasoning output

HuggingFaceDeepseekV32

DeepSeek V3.2 via Hugging Face router

HuggingFaceOpenAIGptOss20b

OpenAI GPT-OSS 20B via Hugging Face router

HuggingFaceOpenAIGptOss120b

OpenAI GPT-OSS 120B via Hugging Face router

HuggingFaceDeepseekV32Novita

DeepSeek V3.2 via Novita on Hugging Face router

HuggingFaceXiaomiMimoV2FlashNovita

Xiaomi MiMo-V2-Flash via Novita on Hugging Face router

HuggingFaceMinimaxM25Novita

MiniMax M2.5 via Novita on Hugging Face router

HuggingFaceGlm5Novita

Z.AI GLM-5 via Novita on Hugging Face router

HuggingFaceQwen3CoderNextNovita

Qwen3-Coder-Next via Novita inference provider on Hugging Face router

XaiGrok4

Grok-4 - Flagship xAI model with advanced reasoning

XaiGrok4Mini

Grok-4 Mini - Efficient xAI model variant

XaiGrok4Code

Grok-4 Code - Code-focused Grok deployment

XaiGrok4CodeLatest

Grok-4 Code Latest - Latest Grok code model with enhanced reasoning tools

XaiGrok4Vision

Grok-4 Vision - Multimodal Grok model

ZaiGlm5

GLM-5 - Flagship Z.ai foundation model for complex systems

MoonshotMinimaxM25

MiniMax-M2.5 - MiniMax model served via Moonshot API

MoonshotQwen3CoderNext

Qwen3-Coder-Next - Qwen3 Coder Next model served via Moonshot API

OllamaGptOss20b

GPT-OSS 20B - Open-weight GPT-OSS 20B model served via Ollama locally

OllamaGptOss20bCloud

GPT-OSS 20B Cloud - Cloud-hosted GPT-OSS 20B served via Ollama Cloud

OllamaGptOss120bCloud

GPT-OSS 120B Cloud - Cloud-hosted GPT-OSS 120B served via Ollama Cloud

OllamaQwen317b

Qwen3 1.7B - Qwen3 1.7B model served via Ollama

OllamaDeepseekV32Cloud

DeepSeek V3.2 Cloud - DeepSeek V3.2 reasoning deployment via Ollama Cloud

OllamaQwen3Next80bCloud

Qwen3 Next 80B Cloud - Next-generation Qwen3 80B via Ollama Cloud

OllamaMistralLarge3675bCloud

Mistral Large 3 675B Cloud - Mistral Large 3 reasoning model via Ollama Cloud

OllamaQwen3Coder480bCloud

Qwen3 Coder 480B Cloud - Cloud-hosted Qwen3 Coder model served via Ollama Cloud

OllamaGemini3ProPreviewLatestCloud

Gemini 3 Pro Preview Latest Cloud - Google Gemini 3 Pro Preview via Ollama Cloud

OllamaDevstral2123bCloud

Devstral 2 123B Cloud - Mistral Devstral 2 123B model via Ollama Cloud

OllamaMinimaxM2Cloud

MiniMax-M2 Cloud - Cloud-hosted MiniMax-M2 model served via Ollama Cloud

OllamaGlm5Cloud

GLM-5 Cloud - Cloud-hosted GLM-5 model served via Ollama Cloud

OllamaMinimaxM25Cloud

MiniMax-M2.5 Cloud - Cloud-hosted MiniMax-M2.5 model served via Ollama Cloud

OllamaGemini3FlashPreviewCloud

Gemini 3 Flash Preview Cloud - Google Gemini 3 Flash Preview via Ollama Cloud

OllamaNemotron3Nano30bCloud

Nemotron-3-Nano 30B Cloud - NVIDIA Nemotron-3-Nano 30B via Ollama Cloud

MinimaxM25

MiniMax-M2.5 - Latest MiniMax model with further improvements in reasoning and coding

MinimaxM2

MiniMax-M2 - MiniMax reasoning-focused model

LmStudioMetaLlama38BInstruct

Meta Llama 3 8B Instruct served locally via LM Studio

LmStudioMetaLlama318BInstruct

Meta Llama 3.1 8B Instruct served locally via LM Studio

LmStudioQwen257BInstruct

Qwen2.5 7B Instruct served locally via LM Studio

LmStudioGemma22BIt

Gemma 2 2B IT served locally via LM Studio

LmStudioGemma29BIt

Gemma 2 9B IT served locally via LM Studio

LmStudioPhi31Mini4kInstruct

Phi-3.1 Mini 4K Instruct served locally via LM Studio

OpenRouterGrokCodeFast1

Grok Code Fast 1 - Fast OpenRouter coding model powered by xAI Grok

OpenRouterGrok4Fast

Grok 4 Fast - Reasoning-focused Grok endpoint with transparent traces

OpenRouterGrok41Fast

Grok 4.1 Fast - Enhanced Grok 4.1 fast inference with improved reasoning

OpenRouterGrok4

Grok 4 - Flagship Grok 4 endpoint exposed through OpenRouter

OpenRouterQwen3Max

Qwen3 Max - Flagship Qwen3 mixture for general reasoning

OpenRouterQwen3235bA22b

Qwen3 235B A22B - Mixture-of-experts Qwen3 235B general model

OpenRouterQwen3235bA22b2507

Qwen3 235B A22B Instruct 2507 - Instruction-tuned Qwen3 235B A22B

OpenRouterQwen3235bA22bThinking2507

Qwen3 235B A22B Thinking 2507 - Deliberative Qwen3 235B A22B reasoning release

OpenRouterQwen332b

Qwen3 32B - Dense 32B Qwen3 deployment

OpenRouterQwen330bA3b

Qwen3 30B A3B - Active-parameter 30B Qwen3 model

OpenRouterQwen330bA3bInstruct2507

Qwen3 30B A3B Instruct 2507 - Instruction-tuned Qwen3 30B A3B

OpenRouterQwen330bA3bThinking2507

Qwen3 30B A3B Thinking 2507 - Deliberative Qwen3 30B A3B release

OpenRouterQwen314b

Qwen3 14B - Lightweight Qwen3 14B model

OpenRouterQwen38b

Qwen3 8B - Compact Qwen3 8B deployment

OpenRouterQwen3Next80bA3bInstruct

Qwen3 Next 80B A3B Instruct - Next-generation Qwen3 instruction model

OpenRouterQwen3Next80bA3bThinking

Qwen3 Next 80B A3B Thinking - Next-generation Qwen3 reasoning release

OpenRouterQwen35Plus0215

Qwen3.5-397B-A17B - Native vision-language model with linear attention and sparse MoE, 1M context window

OpenRouterQwen3Coder

Qwen3 Coder - Qwen3-based coding model tuned for IDE workflows

OpenRouterQwen3CoderPlus

Qwen3 Coder Plus - Premium Qwen3 coding model with long context

OpenRouterQwen3CoderFlash

Qwen3 Coder Flash - Latency optimised Qwen3 coding model

OpenRouterQwen3Coder30bA3bInstruct

Qwen3 Coder 30B A3B Instruct - Large Mixture-of-Experts coding deployment

OpenRouterQwen3CoderNext

Qwen3 Coder Next - Next-generation Qwen3 coding model with enhanced reasoning

OpenRouterDeepseekChat

DeepSeek V3.2 Chat - Official chat model via OpenRouter

OpenRouterDeepSeekV32

DeepSeek V3.2 - Standard model with thinking support via OpenRouter

OpenRouterDeepseekReasoner

DeepSeek V3.2 Reasoner - Thinking mode via OpenRouter

OpenRouterDeepSeekV32Speciale

DeepSeek V3.2 Speciale - Enhanced reasoning model (no tool-use)

OpenRouterDeepSeekV32Exp

DeepSeek V3.2 Exp - Experimental DeepSeek V3.2 listing

OpenRouterDeepSeekChatV31

DeepSeek Chat v3.1 - Advanced DeepSeek model via OpenRouter

OpenRouterDeepSeekR1

DeepSeek R1 - DeepSeek R1 reasoning model with chain-of-thought

OpenRouterOpenAIGptOss120b

OpenAI gpt-oss-120b - Open-weight 120B reasoning model via OpenRouter

OpenRouterOpenAIGptOss120bFree

OpenAI gpt-oss-120b:free - Open-weight 120B reasoning model free tier via OpenRouter

OpenRouterOpenAIGptOss20b

OpenAI gpt-oss-20b - Open-weight 20B deployment via OpenRouter

OpenRouterOpenAIGpt5

OpenAI GPT-5 - OpenAI GPT-5 model accessed through OpenRouter

OpenRouterOpenAIGpt5Codex

OpenAI GPT-5 Codex - OpenRouter listing for GPT-5 Codex

OpenRouterOpenAIGpt5Chat

OpenAI GPT-5 Chat - Chat optimised GPT-5 endpoint without tool use

OpenRouterAnthropicClaudeSonnet45

Claude Sonnet 4.5 - Anthropic Claude Sonnet 4.5 listing

OpenRouterAnthropicClaudeHaiku45

Claude Haiku 4.5 - Anthropic Claude Haiku 4.5 listing

OpenRouterAnthropicClaudeOpus41

Claude Opus 4.1 - Anthropic Claude Opus 4.1 listing

OpenRouterAmazonNova2LiteV1

Amazon Nova 2 Lite - Amazon Nova 2 Lite model via OpenRouter

OpenRouterMistralaiMistralLarge2512

Mistral Large 3 2512 - Mistral Large 3 2512 model via OpenRouter

OpenRouterNexAgiDeepseekV31NexN1

DeepSeek V3.1 Nex N1 - Nex AGI DeepSeek V3.1 Nex N1 model via OpenRouter

OpenRouterOpenAIGpt51

OpenAI GPT-5.1 - OpenAI GPT-5.1 model accessed through OpenRouter

OpenRouterOpenAIGpt51Codex

OpenAI GPT-5.1-Codex - OpenRouter listing for GPT-5.1 Codex

OpenRouterOpenAIGpt51CodexMax

OpenAI GPT-5.1-Codex-Max - OpenRouter listing for GPT-5.1 Codex Max

OpenRouterOpenAIGpt51CodexMini

OpenAI GPT-5.1-Codex-Mini - OpenRouter listing for GPT-5.1 Codex Mini

OpenRouterOpenAIGpt51Chat

OpenAI GPT-5.1 Chat - Chat optimised GPT-5.1 endpoint without tool use

OpenRouterOpenAIGpt52

OpenAI GPT-5.2 - OpenAI GPT-5.2 model accessed through OpenRouter

OpenRouterOpenAIGpt52Chat

OpenAI GPT-5.2 Chat - Chat optimised GPT-5.2 endpoint without tool use

OpenRouterOpenAIGpt52Codex

OpenAI GPT-5.2-Codex - OpenRouter listing for GPT-5.2 Codex

OpenRouterOpenAIGpt52Pro

OpenAI GPT-5.2 Pro - Professional tier GPT-5.2 model accessed through OpenRouter

OpenRouterOpenAIO1Pro

OpenAI o1-pro - OpenAI o1-pro advanced reasoning model via OpenRouter

OpenRouterStepfunStep35FlashFree

Step 3.5 Flash (free) - StepFun’s most capable open-source reasoning model via OpenRouter

OpenRouterZaiGlm5

GLM-5 - Z.AI GLM-5 flagship foundation model via OpenRouter

OpenRouterMoonshotaiKimiK20905

MoonshotAI: Kimi K2 0905 - MoonshotAI Kimi K2 0905 MoE release optimised for coding agents

OpenRouterMoonshotaiKimiK2Thinking

MoonshotAI: Kimi K2 Thinking - MoonshotAI reasoning-tier Kimi K2 release optimized for long-horizon agents

OpenRouterMoonshotaiKimiK25

MoonshotAI: Kimi K2.5 - MoonshotAI Kimi K2.5 multimodal model with long-context and reasoning capabilities via OpenRouter

OpenRouterMinimaxM25

MiniMax-M2.5 - MiniMax flagship model via OpenRouter

Implementations

impl ModelId

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

Convert the model identifier to its string representation used in API calls and configurations

impl ModelId

pub fn is_flash_variant(&self) -> bool

Check if this is a “flash” variant (optimized for speed)

pub fn is_pro_variant(&self) -> bool

Check if this is a “pro” variant (optimized for capability)

pub fn is_efficient_variant(&self) -> bool

Check if this is an optimized/efficient variant

pub fn is_top_tier(&self) -> bool

Check if this is a top-tier model

pub fn is_reasoning_variant(&self) -> bool

Determine whether the model is a reasoning-capable variant

pub fn supports_tool_calls(&self) -> bool

Determine whether the model supports tool calls/function execution

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

Get the generation/version string for this model

impl ModelId

pub fn openrouter_vendor(&self) -> Option<&'static str>

Return the OpenRouter vendor slug when this identifier maps to a marketplace listing

pub fn all_models() -> Vec<ModelId>

Get all available models as a vector

pub fn models_for_provider(provider: Provider) -> Vec<ModelId>

Get all models for a specific provider

impl ModelId

pub fn fallback_models() -> Vec<ModelId>

Get recommended fallback models in order of preference

pub fn default_orchestrator() -> Self

Get the default orchestrator model (more capable)

pub fn default_subagent() -> Self

Get the default subagent model (fast and efficient)

pub fn default_orchestrator_for_provider(provider: Provider) -> Self

Get provider-specific defaults for orchestrator

pub fn default_subagent_for_provider(provider: Provider) -> Self

Get provider-specific defaults for subagent

pub fn default_single_for_provider(provider: Provider) -> Self

Get provider-specific defaults for single agent

impl ModelId

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

Get a description of the model’s characteristics

impl ModelId

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

Get the display name for the model (human-readable)

impl ModelId

pub fn provider(&self) -> Provider

Get the provider for this model

pub fn supports_reasoning_effort(&self) -> bool

Whether this model supports configurable reasoning effort levels

Trait Implementations

impl Clone for ModelId

fn clone(&self) -> ModelId

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ModelId

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

Formats the value using the given formatter.

impl Default for ModelId

fn default() -> ModelId

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for ModelId

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Display for ModelId

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

Formats the value using the given formatter.

impl FromStr for ModelId

type Err = ModelParseError

The associated error which can be returned from parsing.

fn from_str(s: &str) -> Result<Self, Self::Err>

Parses a string s to return a value of this type.

impl Hash for ModelId

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl JsonSchema for ModelId

fn schema_name() -> Cow<'static, str>

The name of the generated JSON Schema.

fn schema_id() -> Cow<'static, str>

Returns a string that uniquely identifies the schema produced by this type.

fn json_schema(generator: &mut SchemaGenerator) -> Schema

Generates a JSON Schema for this type.

fn inline_schema() -> bool

Whether JSON Schemas generated for this type should be included directly in parent schemas, rather than being re-used where possible using the $ref keyword.

impl PartialEq for ModelId

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

Tests for self and other values to be equal, and is used by ==.

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

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for ModelId

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

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Copy for ModelId

impl Eq for ModelId

impl StructuralPartialEq for ModelId