--- name: zg8qqc21.nx.link description: zg8qqc21.nx.link exposes four read-only discovery endpoints for Venice.ai's model and image-style catalog. It allows agents to enumerate available models, resolve capability-based trait labels, translate third-party model identifiers to Venice.ai equivalents, and list supported image-generation styles. host: zg8qqc21.nx.link --- # zg8qqc21.nx.link This host serves as a pre-inference discovery layer for Venice.ai. It is aimed at agents that need to programmatically select the right Venice.ai model or image style before submitting a generation request. All four skills are catalog/metadata lookups with no inputs required; none of them perform inference, generate content, or modify state. ## When to use this host Use this host when an agent needs to discover or select Venice.ai models or image styles before making an inference or generation call — particularly when translating from OpenAI or Anthropic model names, resolving capability-based model selection, or presenting style options to a user. Do not use this host to actually run inference, generate images, or check real-time model availability or status; all skills return static or near-static catalog data. For live inference on Venice.ai, route to a host that exposes Venice.ai's chat completions or image generation endpoints directly. For model discovery on other platforms (OpenAI, Anthropic, etc.), use those platforms' own catalog APIs. ## Capabilities ### Model Discovery and Selection These three skills together allow an agent to fully enumerate and identify Venice.ai models by full catalog, by capability trait, or by mapping from a known third-party identifier. - **`fetch-venice-model-catalog`** — Returns a full list of Venice.ai models with metadata including context length, pricing, capabilities, and model type for discovery and selection. - **`fetch-venice-model-traits`** — Returns a mapping of Venice.ai model trait labels (default, most_intelligent, most_uncensored, default_reasoning, default_vision, default_code) to their current model identifiers. - **`fetch-model-compatibility-mapping`** — Returns a mapping of third-party AI model identifiers (OpenAI, Anthropic, etc.) to their Venice.ai equivalent model names. ### Image Style Discovery Returns the list of available image-generation style names on Venice.ai, enabling agents or users to select a valid style before submitting an image request. - **`fetch-image-styles`** — Returns the full list of available image-generation style names from Venice.ai, including artistic, photographic, game-themed, and paper-craft categories. ## Workflows ### Translate External Model and Confirm Capabilities *Use when an agent needs to migrate a prompt pipeline from a third-party model (e.g. gpt-4o, claude-3-5-sonnet) to Venice.ai and wants to verify the mapped model's context length, pricing, and capabilities before use.* 1. **`fetch-model-compatibility-mapping`** — Resolve the third-party model name to its Venice.ai equivalent identifier. 2. **`fetch-venice-model-catalog`** — Look up the resolved Venice.ai model ID in the full catalog to confirm its context length, pricing tier, and supported capabilities such as vision or function calling. ### Select Best Model for a Capability *Use when an agent needs to pick the optimal Venice.ai model for a specific capability (reasoning, vision, code, uncensored output) and then verify its full specs before committing.* 1. **`fetch-venice-model-traits`** — Retrieve the trait-to-model-ID mapping to identify the recommended model for the target capability. 2. **`fetch-venice-model-catalog`** — Cross-reference the returned model ID against the full catalog to confirm pricing, context window, and any additional capability flags. ## Skill reference ### `fetch-image-styles` **Venice.ai Image Styles** — Returns the full list of available image-generation style names from Venice.ai, including artistic, photographic, game-themed, and paper-craft categories. *Use when:* Use when an agent or user needs to know which image styles are supported before submitting an image-generation request, or to present style options to a user for selection. *Not for:* Do not use to generate images or retrieve style metadata beyond names; this endpoint returns style names only, not descriptions, previews, or configuration parameters. **Returns:** Returns an object with a 'data' array of ~80 image style name strings (e.g. 'Anime', 'Photographic', 'Pixel Art', 'Watercolor') and an 'object' field set to 'list'. **Example:** `GET https://zg8qqc21.nx.link/image/styles` --- ### `fetch-venice-model-catalog` **Venice AI Model Catalog** — Returns a full list of Venice.ai models with metadata including context length, pricing, capabilities, and model type for discovery and selection. *Use when:* Use when an agent needs to discover available Venice.ai models, compare their capabilities (vision, function calling, reasoning, web search), pricing tiers, or context window sizes before selecting a model for a task. *Not for:* Do not use for querying model availability on other AI platforms; this returns only Venice.ai-hosted models. Not suitable for real-time model status checks — availability may lag actual deployment state. **Returns:** Returns a JSON object with a 'data' array of model entries, each containing id, type, context_length, pricing in USD/diem, and a capabilities object with flags for function calling, vision, reasoning, web search, and more. **Example:** `GET https://zg8qqc21.nx.link/models` --- ### `fetch-venice-model-traits` **Venice Model Traits** — Returns a mapping of Venice.ai model trait labels (default, most_intelligent, most_uncensored, default_reasoning, default_vision, default_code) to their current model identifiers. *Use when:* Use when an agent needs to discover which Venice.ai model to use for a specific capability such as reasoning, vision, code generation, or uncensored output, before making a model inference call. *Not for:* Do not use to list all available models or retrieve full model specs; this only returns trait-to-model-ID mappings, not model capabilities or pricing details. **Returns:** Returns a data object mapping trait labels (default, most_intelligent, function_calling_default, most_uncensored, default_reasoning, default_vision, default_code) to their current Venice.ai model IDs, plus object and type fields. **Example:** `GET https://zg8qqc21.nx.link/models/traits` --- ### `fetch-model-compatibility-mapping` **Model Compatibility Mapper** — Returns a mapping of third-party AI model identifiers (OpenAI, Anthropic, etc.) to their Venice.ai equivalent model names. *Use when:* Use when an agent needs to translate a well-known model name (e.g. gpt-4o, claude-3-5-sonnet) to the corresponding Venice.ai model identifier before making an inference request on Venice.ai. *Not for:* Do not use for real-time model availability or pricing; this endpoint returns static compatibility mappings only, not live model status. **Returns:** Returns a data object mapping third-party model names (gpt-4o, claude-3-5-sonnet, o1-mini, etc.) to Venice.ai model identifiers, plus object='list' and type='text'. **Example:** `GET https://zg8qqc21.nx.link/models/compatibility_mapping` ---