Sarvam Mcp

Organization

sarvamai

sarvam-mcp

Run Indic speech and language tasks through the Sarvam MCP server: translate text, transcribe or dub audio, TTS, localize i18n files, LLM complete, document parse / extract / classify / split, or fetch Sarvam API docs/snippets while coding. Use this skill when the user wants Sarvam to act in chat, when sarvam_tools_* / sarvam_code_* are available, or when setting up sarvam-mcp in Cursor, Claude Code, Claude Desktop, Windsurf, Zed, Codex, Gemini CLI, VS Code, Cline, Continue, or LM Studio — even if they only say "translate this to Hindi" or "make it speak." Do not use for LiveKit/Pipecat agent code (voice-agents) or SDK-only coding when MCP is not installed (translate/text, translate/document, speech-to-text, text-to-speech, chat).

Overview

Publisher	sarvamai
Repository	`skills`
Skill name	`sarvam-mcp`
Stars	72
Forks	24
Bundled files	2
License	Apache-2.0
Links	Skill source Repository

Markdown instructions
A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.
Works with any LLM
AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.
2 bundled files
Scripts, templates, and references the model can read while it works. Files are read-only and never executed.
Open source
Published by sarvamai on GitHub. Read the source before you install it.

Installation

Install the Sarvam Mcp AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
https://github.com/sarvamai/skills/tree/main/sarvam-mcp
Enable the skill in any chat where you want it available.

Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills

git clone --depth 1 https://github.com/sarvamai/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/sarvam-mcp .claude/skills/sarvam-mcp

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Sarvam Mcp in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Sarvam Mcp on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Sarvam Mcp is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

Loaded only when it is needed
The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.
Switch models mid-chat
Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Sarvam MCP

Drive sarvam-mcp. Prefer MCP tools over inventing HTTP/SDK calls for live work.

Procedure

Copy and track:


- [ ] 1. MCP available? If no → read references/install.md (or fall back to SDK skills)
- [ ] 2. Namespace: live action → sarvam_tools_*; write/lookup code → sarvam_code_*
- [ ] 3. Prefer a composite if the task spans STT/translate/TTS/LLM
- [ ] 4. Call the tool (absolute paths; BCP-47 codes)
- [ ] 5. On auth error → sarvam_tools_set_api_key; on param error → fix and retry
- [ ] 6. Return the result path/text; mention observability only if debugging

1. Namespace (mandatory)

User wants…	Namespace	Example
Sarvam to do it now	`sarvam_tools_*`	"Translate this to Tamil"
Help writing code / API facts	`sarvam_code_*`	"How do I call TTS from Python?"

Never use sarvam_code_* for a live translate/TTS/STT request. Never burn sarvam_tools_* credits just to draft code unless the user asked for a live demo.

2. Pick the tool (defaults)

Composites first — do not hand-chain STT → translate → TTS:

Task	Tool
Spoken reply to audio	`sarvam_tools_voice`
Dub into another Indic language	`sarvam_tools_dub`
Localize JSON/YAML/PO/etc.	`sarvam_tools_localize`
Q&A / summary over audio	`sarvam_tools_recall`

Atomic runtime (use when no composite fits):

Task	Tool
Transcribe	`sarvam_tools_stt_transcribe`
Speech → English	`sarvam_tools_stt_translate`
Audio >~30s / diarization	`sarvam_tools_stt_batch_submit` → `_stt_batch_status`
Speak text	`sarvam_tools_tts_speak`
Translate text	`sarvam_tools_translate`
Transliterate / LID / analytics	`sarvam_tools_transliterate` / `_identify_language` / `_text_analytics`
Chat complete	`sarvam_tools_llm_complete`
Document → text / HTML / JSON	`sarvam_tools_vision_parse`
Document fields (schema)	`sarvam_tools_vision_extract`
Label a document or pages	`sarvam_tools_vision_classify`
Split a multi-document PDF	`sarvam_tools_vision_split`
Resume / check a document job	`sarvam_tools_vision_job_status`
Pronunciation dicts	`sarvam_tools_pronunciation_*`
Set / rotate API key	`sarvam_tools_set_api_key`

Build-time (coding help):

Task	Tool
Unsure which model/lang	`sarvam_code_recommend_model`
Endpoint shape	`sarvam_code_api_reference`
Snippet	`sarvam_code_snippet`
Speakers / languages	`sarvam_code_speakers` / `_languages`
Validate draft body	`sarvam_code_validate_request`
Docs search / pricing	`sarvam_code_search_docs` / `_pricing`

Coding flow: recommend_model → snippet or api_reference → validate_request.

For parameters and env knobs, read references/tools.md.

3. Defaults (use unless user overrides)

Setting	Default
STT model	`saaras:v3`
TTS model / speaker	SDK code: `bulbul:v4-flash` / persona ID (`aparna_hi_customer`). Live `tts_speak` / `tts_stream` / `voice` / `dub` (sarvam-mcp ≥ 0.2.11): `model` / `tts_model` = `bulbul:v4-flash` + persona ID; omit the speaker for the `shubh_enhi_ads` default. On older servers the schema enums only `bulbul:v3` + short names (`priya`) — check the connected schema and do not send a v4 persona ID unless it lists `bulbul:v4-flash`.
LLM	`sarvam-30b` (use `sarvam-105b` for hard reasoning)
Translate model	`mayura:v1` (switch to `sarvam-translate:v1` for broader Indic coverage)
Audio path	Absolute local path
Language codes	BCP-47 (`hi-IN`, `ta-IN`, `od-IN` not `or-IN`)

4. Auth

Key sources (env wins): SARVAM_API_KEY, or ~/.sarvam/credentials (api_key = sk_...).

Dashboard: https://dashboard.sarvam.ai/key-management

On auth failure:

sarvam_tools_set_api_key with empty api_key → instructions + link
Call again with sk_... → persists to ~/.sarvam/credentials

Wire auth is api-subscription-key — never invent Authorization: Bearer.

Examples

Live translate

User: "Translate 'Good morning' to Hindi."

→ sarvam_tools_translate with input, source_language_code=en-IN, target_language_code=hi-IN. Return translated_text.

Live TTS

User: "Say नमस्ते in Hindi."

→ sarvam_tools_tts_speak with native-script text, target_language_code=hi-IN. Prefer model=bulbul:v4-flash with a persona ID (aparna_hi_customer) when the tool schema lists it (sarvam-mcp ≥ 0.2.11); on older servers use speaker=priya on bulbul:v3. Return the audio path.

Dub

User: "Dub this clip into Tamil" + path.

→ sarvam_tools_dub with audio_path, target_language_code=ta-IN — not STT + translate + TTS separately.

Code help

User: "Show a Python TTS example."

→ sarvam_code_snippet with api=tts, language=python (not sarvam_tools_tts_speak unless they want audio now).

Gotchas

Mistake agents make	Correct behavior
Invent `curl`/SDK for a live ask	Call `sarvam_tools_*` when MCP is connected
Hand-chain STT→translate→TTS	Use `voice` / `dub` / `localize` / `recall`
REST STT on long files	Use `stt_batch_*` above ~30s
TTS target outside ~11 langs	Check with `sarvam_code_languages` (`api=tts`); STT has ~23
v2 speaker on Bulbul v3	Use `priya`/`shubh` or `sarvam_code_speakers`
v3 short name on `bulbul:v4-flash` (or a v4 persona on v3)	Speaker/model mismatch is rejected. v4-flash takes persona IDs only (`aparna_hi_customer`); v3 takes `shubh` / `priya`. Check `sarvam_code_speakers` with `model=`.
`pitch` / `loudness`	`pitch` −0.5–0.5, `loudness` 0.1–2.5, `pace` 0.5–2.0; out-of-range values are rejected. Accepted on both v3 and v4-flash (live-verified 2026-10-06).
Romanized Indic for TTS	Prefer native script
`output_script` on `sarvam-translate:v1`	`mayura:v1` only
Relative audio paths	Prefer absolute paths; or `audio_base64`/`audio_url` + `filename`
Assume output cwd	Files go under `SARVAM_MCP_BASE_PATH` (default `~/Desktop`) unless tool returns another path
Stale tool names without `_tools`/`_code`	Match the connected server's tool list

When MCP is missing

Read references/install.md and help the user connect the server.
If they only need code: use sibling skills translate/text, translate/document, speech-to-text, text-to-speech, chat, or https://docs.sarvam.ai/llms.txt.

References (load on demand)

File	Read when
references/install.md	MCP not connected, setup/troubleshoot, or user names a client
references/tools.md	Choosing parameters, env vars, or an uncommon tool

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

references/install.md4 KB
references/tools.md5 KB

Frequently asked questions

What does the Sarvam Mcp AI skill do?

Why use Sarvam Mcp on TypingMind?

Because you install it once and use it with any model. Sarvam Mcp is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Sarvam Mcp in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/sarvamai/skills/tree/main/sarvam-mcp. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Sarvam Mcp?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Sarvam Mcp?

As many as you like. As long as a model supports skills, you can use Sarvam Mcp with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Sarvam Mcp AI skill free?

Yes. It is published on GitHub by sarvamai under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Openclaw Testing

openclaw

OrganizationPopular

Choose proportional OpenClaw tests and checks, diagnose failures, and route environment-sensitive or release proof to its owner.

Release Openclaw Ci

Run, watch, debug, and summarize OpenClaw full release CI, release checks, live provider gates, install/update proofs, and release-secret preflights.

Release Openclaw Mac

Run or recover OpenClaw macOS release signing, notarization, appcast, and asset promotion.

Release Openclaw Maintainer

openclaw

OrganizationPopular

Prepare, publish, recover, or verify OpenClaw beta, stable, and extended-stable releases, including approved backports.

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Sarvam MCP

Procedure

1. Namespace (mandatory)

2. Pick the tool (defaults)

3. Defaults (use unless user overrides)

4. Auth

Examples

Gotchas

When MCP is missing

References (load on demand)

Bundled files

Frequently asked questions

Related AI Skills

Set up your own AI workspace now