vtcode_config/models/
model_id.rs

1use serde::{Deserialize, Serialize};
2
3mod as_str;
4mod capabilities;
5mod collection;
6mod defaults;
7mod description;
8mod display;
9mod format;
10mod openrouter;
11mod parse;
12mod provider;
13
14pub use capabilities::{
15    ModelCatalogEntry, ModelPricing, catalog_provider_keys, model_catalog_entry,
16    supported_models_for_provider,
17};
18
19/// Centralized enum for all supported model identifiers
20#[cfg_attr(feature = "schema", derive(schemars::JsonSchema))]
21#[derive(Clone, Copy, Debug, Default, PartialEq, Eq, Hash, Serialize, Deserialize)]
22pub enum ModelId {
23    // Gemini models
24    /// Gemini 3.1 Pro Preview - Latest Gemini 3.1 Pro flagship
25    Gemini31ProPreview,
26    /// Gemini 3.1 Pro Preview Custom Tools - Optimized for custom tools & bash
27    Gemini31ProPreviewCustomTools,
28    /// Gemini 3.1 Flash Lite Preview - Most cost-efficient model, offering fastest performance for high-frequency, lightweight tasks
29    Gemini31FlashLitePreview,
30    /// Gemini 3.5 Flash - High-efficiency frontier model for fast inference
31    #[default]
32    Gemini35Flash,
33
34    // OpenAI models
35    /// GPT-5.5 - Next-gen OpenAI model dated release (2026-04-23)
36    GPT55,
37    /// GPT-5.4 - Mainline frontier GPT model for general-purpose and coding work
38    GPT54,
39    /// GPT-5.4 Pro - Higher-compute GPT-5.4 variant for difficult problems
40    GPT54Pro,
41    /// GPT-5.4 Nano - Lightweight GPT-5.4 variant optimized for speed and cost-efficiency
42    GPT54Nano,
43    /// GPT-5.4 Mini - Compact GPT-5.4 variant for cost-effective tasks with reduced reasoning overhead
44    GPT54Mini,
45    /// GPT-5.3 Codex - Code-focused GPT-5.3 variant optimized for agentic coding with reasoning effort support (low, medium, high, xhigh)
46    GPT53Codex,
47    /// GPT-OSS 20B - OpenAI's open-source 20B parameter model using harmony
48    OpenAIGptOss20b,
49    /// GPT-OSS 120B - OpenAI's open-source 120B parameter model using harmony
50    OpenAIGptOss120b,
51
52    // Anthropic models
53    /// Claude Opus 4.8 - Anthropic's most capable model for complex reasoning and agentic coding
54    ClaudeOpus48,
55    /// Claude Sonnet 4.6 - Balanced flagship Anthropic model in VT Code's conservative rollout
56    ClaudeSonnet46,
57    /// Claude Haiku 4.5 - Latest efficient Anthropic model (2025-10-15)
58    ClaudeHaiku45,
59    /// Claude Mythos Preview - Invitation-only Anthropic research preview for defensive cybersecurity workflows
60    ClaudeMythosPreview,
61    /// GitHub Copilot auto model selection
62    CopilotAuto,
63    /// GitHub Copilot GPT-5.2 Codex
64    CopilotGPT52Codex,
65    /// GitHub Copilot GPT-5.1 Codex Max
66    CopilotGPT51CodexMax,
67    /// GitHub Copilot GPT-5.4
68    CopilotGPT54,
69    /// GitHub Copilot GPT-5.4 Mini
70    CopilotGPT54Mini,
71    /// GitHub Copilot Claude Sonnet 4.6
72    CopilotClaudeSonnet46,
73
74    // DeepSeek models
75    /// DeepSeek V4 Pro - High-performance reasoning model with advanced thinking
76    DeepSeekV4Pro,
77    /// DeepSeek V4 Flash - Fast inference model for cost-effective reasoning
78    DeepSeekV4Flash,
79
80    // Mistral AI models
81    /// Mistral Large 3 - State-of-the-art open-weight general-purpose multimodal model
82    MistralLarge3,
83    // Hugging Face models
84    /// OpenAI GPT-OSS 20B via Hugging Face router
85    HuggingFaceOpenAIGptOss20b,
86    /// OpenAI GPT-OSS 120B via Hugging Face router
87    HuggingFaceOpenAIGptOss120b,
88    /// MiniMax M2.5 via Novita on Hugging Face router
89    HuggingFaceMinimaxM25Novita,
90    /// Z.AI GLM-5 via Novita on Hugging Face router
91    HuggingFaceGlm5Novita,
92    /// Z.AI GLM-5.1 via zai-org provider on Hugging Face router
93    HuggingFaceGlm51ZaiOrg,
94    /// Qwen3-Coder-Next via Novita inference provider on Hugging Face router
95    HuggingFaceQwen3CoderNextNovita,
96    /// Qwen3.5-397B-A17B via Together inference provider on Hugging Face router
97    HuggingFaceQwen35397BA17BTogether,
98    /// Kimi K2.6 via Novita on Hugging Face router
99    HuggingFaceKimiK26Novita,
100    /// DeepSeek V4 Flash via Novita on Hugging Face router
101    HuggingFaceDeepseekV4FlashNovita,
102    /// DeepSeek V4 Pro via Together on Hugging Face router
103    HuggingFaceDeepseekV4ProTogether,
104    /// Step 3.5 Flash via Hugging Face router
105    HuggingFaceStep35Flash,
106    /// Z.AI GLM-5.1 via DeepInfra inference provider on Hugging Face router
107    HuggingFaceGlm51Deepinfra,
108    /// MiniMax M2.7 via Novita on Hugging Face router
109    HuggingFaceMinimaxM27Novita,
110    /// DeepSeek V4 Pro via Novita on Hugging Face router
111    HuggingFaceDeepseekV4ProNovita,
112    /// NVIDIA Nemotron 3 Ultra 550B via Together on Hugging Face router
113    HuggingFaceNvidiaNemotron3Ultra550bA55bNvfp4Together,
114
115    // StepFun models
116    /// Step 3.7 Flash - StepFun's flagship multimodal reasoning model with tool calling
117    StepFun37Flash,
118
119    // Evolink gateway models (namespaced as `evolink/<model>`)
120    /// GPT-5.2 served through the Evolink gateway
121    EvolinkGpt52,
122    /// GPT-5.5 served through the Evolink gateway
123    EvolinkGpt55,
124    /// DeepSeek V4 Pro served through the Evolink gateway
125    EvolinkDeepseekV4Pro,
126    /// DeepSeek V4 Flash served through the Evolink gateway
127    EvolinkDeepseekV4Flash,
128    /// Doubao Seed 2.0 Pro served through the Evolink gateway
129    EvolinkDoubaoSeed20Pro,
130    /// Gemini 3.1 Pro served through the Evolink gateway (OpenAI SDK format)
131    EvolinkGemini31Pro,
132    /// Gemini 3.5 Flash served through the Evolink gateway (OpenAI SDK format)
133    EvolinkGemini35Flash,
134    /// MiniMax-M3 served through the Evolink gateway (OpenAI Chat Completions format)
135    EvolinkMinimaxM3,
136    /// Claude Sonnet 4.6 served through the Evolink gateway (Anthropic Messages API)
137    EvolinkClaudeSonnet46,
138    /// Claude Opus 4.8 served through the Evolink gateway (Anthropic Messages API)
139    EvolinkClaudeOpus48,
140    /// Claude Haiku 4.5 served through the Evolink gateway (Anthropic Messages API)
141    EvolinkClaudeHaiku45,
142
143    /// GLM-5 - Flagship Z.ai foundation model for complex systems
144    ZaiGlm5,
145    /// GLM-5.1 - Next-gen Z.ai foundation model with improved reasoning
146    ZaiGlm51,
147
148    // MiMo models
149    /// MiMo V2.5 Pro - Xiaomi's flagship reasoning model with 1M context
150    MiMoV25Pro,
151    /// MiMo V2.5 - Xiaomi's omni-modal model with full-modal understanding and 1M context
152    MiMoV25,
153
154    // Moonshot models
155    /// Kimi K2.6 - Moonshot.ai's latest 1T MoE flagship (32B active, MLA, MoonViT vision)
156    MoonshotKimiK26,
157
158    // OpenCode Zen models
159    /// GPT-5.4 - OpenCode Zen default flagship model
160    OpenCodeZenGPT54,
161    /// GPT-5.4 Mini - Lower-cost OpenCode Zen GPT option
162    OpenCodeZenGPT54Mini,
163    /// Claude Sonnet 4.6 - Anthropic-backed OpenCode Zen coding model
164    OpenCodeZenClaudeSonnet46,
165    /// GLM-5.1 - Z.AI model served through OpenCode Zen
166    OpenCodeZenGlm51,
167
168    // OpenCode Go models
169    /// GLM-5.1 - Z.AI model included with OpenCode Go
170    OpenCodeGoGlm51,
171    /// MiniMax M2.5 - OpenCode Go subscription model
172    OpenCodeGoMinimaxM25,
173    /// MiniMax M2.7 - Higher-tier OpenCode Go subscription model
174    OpenCodeGoMinimaxM27,
175
176    // Qwen models
177    /// Qwen 3.7 Max - Alibaba Cloud's flagship reasoning model with 131K context
178    Qwen37Max,
179    /// Qwen 3.6 Flash - Alibaba Cloud's fast inference model with 1M context
180    Qwen36Flash,
181    /// Qwen 3.6 Plus - Alibaba Cloud's balanced model with 131K context
182    Qwen36Plus,
183    /// DeepSeek V4 Flash via Qwen Cloud API
184    QwenDeepSeekV4Flash,
185    /// DeepSeek V4 Pro via Qwen Cloud API
186    QwenDeepSeekV4Pro,
187    /// GLM-5.1 via Qwen Cloud API
188    QwenGlm51,
189
190    // Ollama models
191    /// GPT-OSS 20B - Open-weight GPT-OSS 20B model served via Ollama locally
192    OllamaGptOss20b,
193    /// GPT-OSS 20B Cloud - Cloud-hosted GPT-OSS 20B served via Ollama Cloud
194    OllamaGptOss20bCloud,
195    /// GPT-OSS 120B Cloud - Cloud-hosted GPT-OSS 120B served via Ollama Cloud
196    OllamaGptOss120bCloud,
197    /// Qwen3 1.7B - Qwen3 1.7B model served via Ollama
198    OllamaQwen317b,
199    /// Qwen3 Coder Next - Coding-optimized Qwen3 Next model served via Ollama locally
200    OllamaQwen3CoderNext,
201    /// DeepSeek V4 Flash Cloud - Fast inference DeepSeek V4 Flash model via Ollama Cloud
202    OllamaDeepseekV4FlashCloud,
203    /// DeepSeek V4 Pro Cloud - High-performance DeepSeek V4 Pro model via Ollama Cloud
204    OllamaDeepseekV4ProCloud,
205    /// Qwen3 Next 80B Cloud - Next-generation Qwen3 80B via Ollama Cloud
206    OllamaQwen3Next80bCloud,
207    /// MiniMax-M2 Cloud - Cloud-hosted MiniMax-M2 model served via Ollama Cloud
208    OllamaMinimaxM2Cloud,
209    /// MiniMax-M2.7 Cloud - Cloud-hosted MiniMax-M2.7 model served via Ollama Cloud
210    OllamaMinimaxM27Cloud,
211    /// MiniMax-M3 Cloud - Cloud-hosted MiniMax-M3 model served via Ollama Cloud
212    OllamaMinimaxM3Cloud,
213    /// GLM-5 Cloud - Cloud-hosted GLM-5 model served via Ollama Cloud
214    OllamaGlm5Cloud,
215    /// GLM-5.1 Cloud - Cloud-hosted GLM-5.1 model served via Ollama Cloud
216    OllamaGlm51Cloud,
217    /// MiniMax-M2.5 Cloud - Cloud-hosted MiniMax-M2.5 model served via Ollama Cloud
218    OllamaMinimaxM25Cloud,
219    /// Gemini 3 Flash Preview Cloud - Google Gemini 3 Flash Preview via Ollama Cloud
220    OllamaGemini3FlashPreviewCloud,
221    /// Kimi K2.6 Cloud - Moonshot Kimi K2.6 via Ollama Cloud
222    OllamaKimiK26Cloud,
223    /// Nemotron 3 Super Cloud - NVIDIA Nemotron 3 Super 120B via Ollama Cloud
224    OllamaNemotron3SuperCloud,
225    /// Laguna XS.2 - Poolside's 33B MoE model (3B activated) for agentic coding via Ollama
226    OllamaLagunaXs2,
227
228    // llama.cpp models
229    /// Qwen 3.6 27B - Dense Qwen 3.6 local model served through llama.cpp
230    LlamaCppQwen3627b,
231    /// Qwen 3.6 35B A3B - MoE Qwen 3.6 local model served through llama.cpp
232    LlamaCppQwen3635bA3b,
233    /// Gemma 4 26B A4B - Desktop Gemma 4 MoE model served through llama.cpp
234    LlamaCppGemma426bA4b,
235    /// Gemma 4 E4B - Tiny-footprint Gemma 4 model served through llama.cpp
236    LlamaCppGemma4E4b,
237    /// GPT-OSS 20B - OpenAI open-weight model served through llama.cpp
238    LlamaCppGptOss20b,
239    /// Step 3.5 Flash - StepFun local model served through llama.cpp
240    LlamaCppStep35Flash,
241
242    // MiniMax models
243    /// MiniMax-M3 - Frontier multimodal coding model with 1M context
244    MinimaxM3,
245    /// MiniMax-M2.7 - Recursive self-improvement flagship with 204.8K context
246    MinimaxM27,
247    /// MiniMax-M2.5 - Latest MiniMax model with further improvements in reasoning and coding
248    MinimaxM25,
249
250    // OpenRouter models
251    /// Qwen3 32B - Dense 32B Qwen3 deployment
252    OpenRouterQwen332b,
253    /// Qwen3 30B A3B - Active-parameter 30B Qwen3 model
254    OpenRouterQwen330bA3b,
255    /// Qwen3 30B A3B Instruct 2507 - Instruction-tuned Qwen3 30B A3B
256    OpenRouterQwen330bA3bInstruct2507,
257    /// Qwen3 30B A3B Thinking 2507 - Deliberative Qwen3 30B A3B release
258    OpenRouterQwen330bA3bThinking2507,
259    /// Qwen3 14B - Lightweight Qwen3 14B model
260    OpenRouterQwen314b,
261    /// Qwen3 8B - Compact Qwen3 8B deployment
262    OpenRouterQwen38b,
263    /// Qwen3 Next 80B A3B Instruct - Next-generation Qwen3 instruction model
264    OpenRouterQwen3Next80bA3bInstruct,
265    /// Qwen3 Next 80B A3B Thinking - Next-generation Qwen3 reasoning release
266    OpenRouterQwen3Next80bA3bThinking,
267    /// Qwen3.5-397B-A17B - Native vision-language model with linear attention and sparse MoE, 1M context window
268    OpenRouterQwen35Plus0215,
269    /// Qwen3 Coder - Qwen3-based coding model tuned for IDE workflows
270    OpenRouterQwen3Coder,
271    /// Qwen3 Coder Plus - Premium Qwen3 coding model with long context
272    OpenRouterQwen3CoderPlus,
273    /// Qwen3 Coder Flash - Latency optimised Qwen3 coding model
274    OpenRouterQwen3CoderFlash,
275    /// Qwen3 Coder 30B A3B Instruct - Large Mixture-of-Experts coding deployment
276    OpenRouterQwen3Coder30bA3bInstruct,
277    /// Qwen3 Coder Next - Next-generation Qwen3 coding model with enhanced reasoning
278    OpenRouterQwen3CoderNext,
279    /// DeepSeek V4 Pro - High-performance reasoning model via OpenRouter
280    OpenRouterDeepSeekV4Pro,
281    /// DeepSeek V4 Flash - Fast inference model via OpenRouter
282    OpenRouterDeepSeekV4Flash,
283    /// DeepSeek R1 - DeepSeek R1 reasoning model with chain-of-thought
284    OpenRouterDeepSeekR1,
285    /// OpenAI gpt-oss-120b - Open-weight 120B reasoning model via OpenRouter
286    OpenRouterOpenAIGptOss120b,
287    /// OpenAI gpt-oss-120b:free - Open-weight 120B reasoning model free tier via OpenRouter
288    OpenRouterOpenAIGptOss120bFree,
289    /// OpenAI gpt-oss-20b - Open-weight 20B deployment via OpenRouter
290    OpenRouterOpenAIGptOss20b,
291    /// OpenAI GPT-5 - OpenAI GPT-5 model accessed through OpenRouter
292    OpenRouterOpenAIGpt5,
293    /// OpenAI GPT-5.5 - OpenAI GPT-5.5 model accessed through OpenRouter
294    OpenRouterOpenAIGpt55,
295    /// OpenAI GPT-5 Chat - Chat optimised GPT-5 endpoint without tool use
296    OpenRouterOpenAIGpt5Chat,
297
298    /// Gemini 3.1 Pro Preview - Google's latest Gemini 3.1 Pro model via OpenRouter
299    OpenRouterGoogleGemini31ProPreview,
300
301    /// Claude Sonnet 4.6 - Anthropic Claude Sonnet 4.6 listing
302    OpenRouterAnthropicClaudeSonnet46,
303    /// Claude Haiku 4.5 - Anthropic Claude Haiku 4.5 listing
304    OpenRouterAnthropicClaudeHaiku45,
305    /// Mistral Large 3 2512 - Mistral Large 3 2512 model via OpenRouter
306    OpenRouterMistralaiMistralLarge2512,
307    /// DeepSeek V3.1 Nex N1 - Nex AGI DeepSeek V3.1 Nex N1 model via OpenRouter
308    OpenRouterNexAgiDeepseekV31NexN1,
309    /// Step 3.5 Flash (free) - StepFun's most capable open-source reasoning model via OpenRouter
310    OpenRouterStepfunStep35FlashFree,
311    /// Nemotron 3 Super (free) - NVIDIA's 120B hybrid MoE model via OpenRouter
312    OpenRouterNvidiaNemotron3Super120bA12bFree,
313    /// Nemotron 3 Ultra - NVIDIA's open frontier-reasoning and orchestration model via OpenRouter
314    OpenRouterNvidiaNemotron3Ultra550bA55b,
315    /// GLM-5 - Z.AI GLM-5 flagship foundation model via OpenRouter
316    OpenRouterZaiGlm5,
317    /// GLM-5.1 - Z.AI GLM-5.1 next-gen foundation model via OpenRouter
318    OpenRouterZaiGlm51,
319    /// MiniMax-M2.5 - MiniMax flagship model via OpenRouter
320    OpenRouterMinimaxM25,
321    /// Kimi K2.6 - Moonshot AI's next-generation multimodal model via OpenRouter
322    OpenRouterMoonshotaiKimiK26,
323    /// Qwen3.7 Max - Alibaba's flagship Qwen3.7 model for coding and agentic workloads via OpenRouter
324    OpenRouterQwenQwen37Max,
325    /// Hy3 Preview - Tencent's high-efficiency MoE model for agentic workflows via OpenRouter
326    OpenRouterTencentHy3Preview,
327    /// Grok Build 0.1 - xAI's fast coding model for agentic software engineering via OpenRouter
328    OpenRouterXAiGrokBuild01,
329    /// MiMo-V2.5 - Xiaomi's omnimodal agentic model for complex software engineering via OpenRouter
330    OpenRouterXiaomiMimoV25,
331    /// MiMo-V2.5-Pro - Xiaomi's flagship agentic model for complex software engineering via OpenRouter
332    OpenRouterXiaomiMimoV25Pro,
333    /// Laguna XS.2 (free) - Poolside's efficient free coding agent model via OpenRouter
334    OpenRouterPoolsideLagunaXs2Free,
335    /// Laguna M.1 (free) - Poolside's flagship free coding agent model via OpenRouter
336    OpenRouterPoolsideLagunaM1Free,
337
338    // Poolside models
339    /// Laguna M.1 - Poolside's flagship MoE coding agent model
340    PoolsideLagunaM1,
341    /// Laguna XS.2 - Poolside's efficient MoE coding agent model
342    PoolsideLagunaXs2,
343}
vtcode_config/models/model_id.rs

vtcode_config/models/
model_id.rs
