Medhā MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": true
} |
| resources | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| venice_chatA | Run an OpenAI-compatible chat completion via Venice's uncensored LLM catalog (Claude, GPT-5, Llama, DeepSeek, Qwen, GLM, Kimi, Venice Uncensored 1.1, etc.). Operator preferences — default: minimax-m3-preview; also try: qwen-3-7-max / venice-uncensored-role-play / gemini-3-1-pro-preview. cheap+long (524K ctx) → 1M ctx → character voice → premium reasoning. Uncensored: NSFW prompts allowed where the model permits. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_responsesA | OpenAI-compatible Responses API. Single-turn or multi-turn with tool support. Operator preferences — default: minimax-m3-preview; also try: qwen-3-7-max. OpenAI Responses API; agentic tool-use loop supported. Uncensored: NSFW prompts allowed where the model permits. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_embeddingsA | Compute embeddings for text input (OpenAI-compatible). Operator preferences — default: text-embedding-3-small; also try: gemini-embedding-001. Operator batch preferred; OpenAI-compatible. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_image_generateA | Generate an image. Supports Flux 2 Pro/Max, Lustify SDXL, Anime (WAI), Qwen Image, GPT Image, Nano Banana Pro and others. Operator preferences — default: flux-2-pro; also try: qwen-image / lustify-sdxl / nano-banana-pro. Operator default is flux-2-pro. Use nano-banana-pro for photoreal, anime-wai for stylization, lustify-sdxl for character work. Uncensored: NSFW prompts allowed where the model permits. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_image_editB | Edit an image with a prompt. Returns base64 PNG. Operator preferences — default: firered-image-edit; also try: qwen-image-edit. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_image_multi_editC | Edit multiple images together with a single prompt (multi-image composition / outpainting). Returns base64 PNG. Operator preferences — default: flux-2-pro; also try: qwen-image. Composition / outpainting; 2-8 input images. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_image_upscaleA | Upscale an image (1-4× scale). Endpoint requires base64 image; this tool fetches the URL and uploads it. Returns base64 PNG. Operator preferences — default: flux-2-pro; also try: nano-banana-pro. 1-4× enhancement + replication control. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_image_remove_bgA | Remove image background; returns a transparent PNG (base64). Operator preferences — default: bria-remove-bg. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_video_generateA | Queue a video generation. Supports Sora 2, Veo 3.1, Kling, Wan, LTX 2, Seedance, Runway Gen-4, and others. Pick a specific id like "veo3.1-fast-text-to-video", "veo3.1-fast-image-to-video", "kling-2.6-pro-text-to-video", "wan-2.6-text-to-video", "seedance-2-0-r2v" etc. Operator preferences — default: ltx-2; also try: sora-2 / veo3.1-fast / kling-2.6-pro. Operator prefers ltx-2 for speed. Sora-2 / Veo3.1 / Kling for premium. Seedance 2.0 r2v for video-to-video. Always quote first via venice_video_quote. Uncensored: NSFW prompts allowed where the model permits. Supports x402 wallet auth (no Venice account needed) and API key. Returns { model, queue_id }; poll with venice_video_status. NOTE: 'duration' is a string enum like '4s' / '6s' / '8s' (model-specific, see model card). |
| venice_video_statusA | Check status of a queued video job. Status enum: PROCESSING, COMPLETED. POST endpoint with body {model, queue_id}. POST /v1/video/retrieve — Venice returns PROCESSING or COMPLETED. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_video_completeA | Mark a completed video as downloaded; deletes server-side media. Cleanup hook; removes server-side media. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_video_transcriptionsB | Transcribe a YouTube video URL. Operator preferences — default: youtube-default. YouTube URL → text transcript. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_ttsB | Convert text to speech. Supports cloned voices + emotion tags ([whispers], [sarcastically], etc.). Operator preferences — default: tts-kokoro; also try: elevenlabs-tts. Voice cloning + emotion tags ([whispers], [sarcastically], [laughs], etc.). Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_asrB | Transcribe audio. Fetches the URL server-side and forwards as multipart/form-data file upload. Operator preferences — default: openai/whisper-large-v3. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_voice_cloneB | Manage TTS voices. Action 'list' returns the static catalog of built-in voices grouped by TTS model (Venice does not expose a list endpoint). Action 'create' clones a voice from a sample audio URL via multipart upload to /v1/audio/voices. List built-in voices OR clone from a sample audio URL. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_music_generateA | Queue music generation. Available models: ace-step-15, elevenlabs-music, minimax-music-v2/v25/v26, stable-audio-25, mmaudio-v2-text-to-audio, elevenlabs-sound-effects-v2. Operator preferences — default: ace-step-15; also try: elevenlabs-music / minimax-music-v2 / minimax-music-v25. Operator default = ace-step-15 (broad genre). Elevenlabs-music for vocal. MMaudio-V2 for stem separation. Uncensored: NSFW prompts allowed where the model permits. Supports x402 wallet auth (no Venice account needed) and API key. Returns { model, queue_id }; poll with venice_music_status. |
| venice_music_statusA | Check status of a queued music job (POST endpoint with body {model, queue_id}). POST /v1/audio/retrieve. Async until COMPLETED. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_music_completeB | Mark a completed music job as downloaded. Cleanup hook. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_web_searchB | Search the web (Firecrawl-backed). Returns ranked results with snippets. Operator preferences — default: firecrawl-default. Firecrawl-backed ranked results. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_web_scrapeA | Scrape one URL into markdown text. Operator preferences — default: firecrawl-default. One URL → markdown text. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_text_parserA | Extract text from a document URL. Fetches the URL server-side and uploads the file as multipart/form-data. Operator preferences — default: firecrawl-default. PDF / DOCX / EPUB / PPTX / XLSX → text. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_crypto_rpcA | Proxy a JSON-RPC call to a supported blockchain network (eth_call, eth_blockNumber, etc.). Networks include "base-mainnet", "ethereum-mainnet", "polygon-mainnet", "arbitrum-mainnet", "optimism-mainnet", and others. List all via GET /api/v1/crypto/rpc/networks. Proxies eth_call / eth_blockNumber to Base, Ethereum, Polygon, Arbitrum, Optimism. Supports x402 wallet auth (no Venice account needed) and API key. |
| venice_list_modelsA | List the live model catalog with capabilities and prices. Live catalog with capability flags + USD pricing. No authentication required. |
| venice_image_stylesA | List image style presets available for venice_image_generate. No authentication required. |
| venice_audio_quoteA | Get a price quote for a music generation BEFORE queuing. Useful for budgeting. Pre-flight price quote for music generation. No authentication required. |
| venice_video_quoteA | Get a price quote for a video generation BEFORE queuing. Always quote-generate before queue when model pricing is unknown. No authentication required. |
| venice_list_charactersB | List public Venice characters. Public Venice characters (api-key only, no x402). API key required — this endpoint does not accept x402 wallet auth. |
| venice_chat_with_characterA | Chat with a Venice character by slug. Note: the character lookup itself is API-key-only, but the chat completion supports x402 — so x402 users may need to fetch character info via API key first. Operator preferences — default: venice-uncensored-role-play; also try: venice-uncensored-1-2. Use character_slug from list_characters. Uncensored: NSFW prompts allowed where the model permits. |
| venice_x402_balanceA | Check the prepaid x402 credit balance for a wallet address. SIWX-ONLY: this endpoint rejects API key auth and requires X-Sign-In-With-X (forwarded from VENICE_SIWX_TOKEN). The wallet in the path must match the SIWX-authenticated wallet. |
| venice_x402_top_up_infoA | Fetch step-1 top-up requirements (network, USDC token address, receiver wallet, min amount). Steps 2 (sign USDC authorization) and 3 (POST signed payment) require a wallet and happen OUTSIDE this MCP server. |
| venice_x402_transactionsA | List recent x402 top-up + debit transactions for a wallet. SIWX-ONLY: rejects API key, requires X-Sign-In-With-X (VENICE_SIWX_TOKEN). The wallet in the path must match the SIWX-authenticated wallet. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| uncensored-research | Set up a research session free of nanny refusals (security, medical, legal, journalism). |
| nsfw-creative-writing | Adult fiction / mature creative writing scaffold. |
| image-style-explorer | Generate the same prompt across multiple styles for comparison. |
| medha_music_video_brief | Generate a 30-60s vertical music video: music + 4-8 image frames + video interpolation per frame + optional TTS narration. Operator-recommended model sequence. |
| medha_podcast_pipeline | Triple-tool podcast: web research → script via chat → TTS narration. Operator-curated defaults. |
| medha_dashboard_poster | Generate a hero poster for a landing page or dashboard: chat-composed prompt → flux-2-pro image → optional upscale. |
| medha_character_dossier | Build a persistent role-play character: persona prompt + reference avatar image + voice clone manifest. Save into a single JSON profile for re-use. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| models | Live catalog with prices and capability flags. Auth-free. |
| styles | Available image style presets for venice_image_generate. |
| voices | Available TTS voices including cloned voices. |
| medha://favorites | Medhā operator-curated defaults — a single canonical JSON document listing the operator’s preferred model per tool + per-domain (reasoning / coding / roleplay / vision / longctx). Fetch this in any session to know what the operator prefers before picking a model. |
TDQS
Scored across 31 tools
Most tools are clearly distinct by domain (chat, image, video, music, web, voice, crypto, x402). However, venice_chat and venice_responses overlap significantly — both are chat completions with nearly identical operator preferences and uncensored behavior. Similarly, venice_video_status/venice_music_status and venice_video_complete/venice_music_complete follow parallel but separate verbs that are easy to confuse across media types. The image_edit vs image_multi_edit distinction is reasonably clear, but chat vs responses and the parallel status/complete pairs introduce ambiguity.
The naming follows a strong venice_<domain>_<action> pattern consistently (venice_image_generate, venice_video_status, venice_music_complete, venice_web_search). The domain prefix is uniformly applied to all 31 tools. Minor inconsistency: some tools use generic verbs (chat, asr, tts) while others use domain+verb, and venice_voice_clone uses 'manage' semantics differently from venice_image_styles. Overall highly predictable but with minor phrasing variations.
31 tools is quite heavy for a single server surface. While the underlying Venice API is genuinely broad (chat, images, video, music, voice, web, crypto, x402 billing), the server bundles what could be several focused MCP servers into one, exceeding the comfortable range. The count feels bloated, especially with 5 different x402/crypto tools and 4 video tools plus 4 music tools that duplicate the lifecycle pattern.
The surface is remarkably comprehensive, covering chat, images (generate/edit/multi-edit/upscale/remove-bg/styles), video (generate/status/complete/quote), music (generate/status/complete/quote), voice (TTS/asr/clone), text parsing, web search/scrape, embeddings, and x402 billing. Minor gaps: text parsing accepts URLs but there's no direct file-upload handler for documents, and there's no image generation status polling since it's synchronous, which is fine. The x402 billing flow stops at step-1 direction without steps 2-3, which is acknowledged as external.