Together AI MCP by usefulapi

by usefulapi.io · remote · no sign-in to list tools · v1.16.3 · published by io.usefulapi

Run Together AI chat, embeddings and images; manage fine-tunes, batches and endpoints.

Deletes · 10

Compare Registry entry

23 tools: 10 can delete or cancel things and 13 only read.

Scanned Oct 21, 2026

Deletes or cancels10Claude Code: deny: see these tools
Can't be undone0Claude Code: ask
Changes things0Claude Code: ask
Only reads13Claude Code: allow: see these tools

No check matched anything in the tool definitions. No finding is not a safety guarantee. How the checks work.

claude mcp add --transport http together-ai https://together-ai.usefulapi.io/mcp

Tools

Open a tool for its reasons and full description. Descriptions are the publisher's own words, shown as they are.

23 tools

Deletes or cancels 10Removes, cancels or revokes something.Claude Code: deny

together_cancel_batchCancel a batch job that has not finished.No findings
Why it is in this group
- The name "together_cancel_batch" says "cancel": a word for removing or ending something.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Cancel a batch job that has not finished. Together: POST /batches/{id}/cancel.
Annotations the server declares
- title: Cancel a batch job
- destructiveHint: true
together_cancel_fine_tuneCancel a running fine-tuning job.No findings
Why it is in this group
- The name "together_cancel_fine_tune" says "cancel": a word for removing or ending something.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Cancel a running fine-tuning job. Cannot be resumed, but a new job can continue from its last checkpoint via from_checkpoint. Together: POST /fine-tunes/{id}/cancel.
Annotations the server declares
- title: Cancel a fine-tuning job
- destructiveHint: true
together_chat_completionRun a chat completion on a Together model (billed per token).No findings
Why it is in this group
- The description's first verb is "run": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Run a chat completion on a Together model (billed per token). Non-streaming. For a dedicated endpoint pass its `<project_slug>/<endpoint_slug>` as the model. Together: POST /chat/completions.
Annotations the server declares
- title: Chat completion
- destructiveHint: true
together_create_batchStart an asynchronous batch job over an uploaded JSONL input file (purpose batch-api), at a discount to real-time inference.No findings
Why it is in this group
- The name "together_create_batch" says "create": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Start an asynchronous batch job over an uploaded JSONL input file (purpose batch-api), at a discount to real-time inference. Together: POST /batches.
Annotations the server declares
- title: Create a batch job
- destructiveHint: true
together_create_embeddingsGenerate vector embeddings for one or more texts (billed per token).No findings
Why it is in this group
- The name "together_create_embeddings" says "create": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Generate vector embeddings for one or more texts (billed per token). Together: POST /embeddings.
Annotations the server declares
- title: Create embeddings
- destructiveHint: true
together_create_endpointDeploy a model on dedicated GPUs.No findings
Why it is in this group
- The name "together_create_endpoint" says "create": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Deploy a model on dedicated GPUs. The endpoint STARTS AUTOMATICALLY and bills per minute of uptime until stopped — set inactive_timeout to auto-stop it, and use together_list_hardware for valid hardware ids. Together: POST /endpoints.
Annotations the server declares
- title: Create a dedicated endpoint
- destructiveHint: true
together_create_fine_tuneStart a fine-tuning job on an uploaded training file (billed per token processed).No findings
Why it is in this group
- The name "together_create_fine_tune" says "create": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Start a fine-tuning job on an uploaded training file (billed per token processed). Stop it with together_cancel_fine_tune. Together: POST /fine-tunes.
Annotations the server declares
- title: Create a fine-tuning job
- destructiveHint: true
together_generate_imageGenerate images from a prompt (billed per image/megapixel).No findings
Why it is in this group
- The name "together_generate_image" says "generate": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Generate images from a prompt (billed per image/megapixel). Returns image URLs by default rather than base64, to keep responses small. Together: POST /images/generations.
Annotations the server declares
- title: Generate an image
- destructiveHint: true
together_start_endpointStart a stopped dedicated endpoint.No findings
Why it is in this group
- The name "together_start_endpoint" says "start": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Start a stopped dedicated endpoint. It bills per minute of uptime until stopped. Reversible with together_stop_endpoint. Together: PATCH /endpoints/{endpointId} with state=STARTED.
Annotations the server declares
- title: Start a dedicated endpoint
- destructiveHint: true
together_stop_endpointStop a running dedicated endpoint, which stops its per-minute billing.No findings
Why it is in this group
- The name "together_stop_endpoint" says "stop": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it
Stop a running dedicated endpoint, which stops its per-minute billing. Requests to it fail until it is started again. Together: PATCH /endpoints/{endpointId} with state=STOPPED.
Annotations the server declares
- title: Stop a dedicated endpoint
- destructiveHint: true

Only reads 13Looks things up and changes nothing.Claude Code: allow

together_get_batchFetch one batch job: status (VALIDATING, IN_PROGRESS, COMPLETED, FAILED, EXPIRED, CANCELLED), progress, and the output_file_id / error_file_id once done.No findings
Why it is in this group
- The name "together_get_batch" says "get": a word for reading.
Description, as the server wrote it
Fetch one batch job: status (VALIDATING, IN_PROGRESS, COMPLETED, FAILED, EXPIRED, CANCELLED), progress, and the output_file_id / error_file_id once done. Together: GET /batches/{id}.
Annotations the server declares
- title: Get one batch job
- readOnlyHint: true
together_get_endpointFetch one dedicated endpoint: state, model, hardware, autoscaling bounds and display name.No findings
Why it is in this group
- The name "together_get_endpoint" says "get": a word for reading.
Description, as the server wrote it
Fetch one dedicated endpoint: state, model, hardware, autoscaling bounds and display name. Together: GET /endpoints/{endpointId}.
Annotations the server declares
- title: Get one endpoint
- readOnlyHint: true
together_get_fileFetch one file's metadata, including its processing_status and validation_report (why a fine-tune training file was rejected).No findings
Why it is in this group
- The name "together_get_file" says "get": a word for reading.
Description, as the server wrote it
Fetch one file's metadata, including its processing_status and validation_report (why a fine-tune training file was rejected). Together: GET /files/{id}.
Annotations the server declares
- title: Get one file
- readOnlyHint: true
together_get_fine_tuneFetch one fine-tuning job: status, progress, hyperparameters, token counts, cost and the output model name.No findings
Why it is in this group
- The name "together_get_fine_tune" says "get": a word for reading.
Description, as the server wrote it
Fetch one fine-tuning job: status, progress, hyperparameters, token counts, cost and the output model name. Together: GET /fine-tunes/{id}.
Annotations the server declares
- title: Get one fine-tuning job
- readOnlyHint: true
together_list_batchesList batch inference jobs with status, progress, model and input/output/error file ids.No findings
Why it is in this group
- The name "together_list_batches" says "list": a word for reading.
Description, as the server wrote it
List batch inference jobs with status, progress, model and input/output/error file ids. Together: GET /batches.
Annotations the server declares
- title: List batch jobs
- readOnlyHint: true
together_list_endpointsList endpoints with model, owner and state (PENDING, STARTING, STARTED, STOPPING, STOPPED, ERROR).No findings
Why it is in this group
- The name "together_list_endpoints" says "list": a word for reading.
Description, as the server wrote it
List endpoints with model, owner and state (PENDING, STARTING, STARTED, STOPPING, STOPPED, ERROR). Use mine=true and type=dedicated to see what is running on your account and billing by the minute. Together: GET /endpoints.
Annotations the server declares
- title: List endpoints
- readOnlyHint: true
together_list_evaluationsList LLM-as-a-judge evaluation jobs (classify, score, compare) with status, parameters and results once completed.No findings
Why it is in this group
- The name "together_list_evaluations" says "list": a word for reading.
Description, as the server wrote it
List LLM-as-a-judge evaluation jobs (classify, score, compare) with status, parameters and results once completed. Together: GET /evaluation.
Annotations the server declares
- title: List evaluation jobs
- readOnlyHint: true
together_list_filesList uploaded data files (fine-tune, eval and batch-api inputs, plus job outputs) with size, type, purpose and validation status.No findings
Why it is in this group
- The name "together_list_files" says "list": a word for reading.
Description, as the server wrote it
List uploaded data files (fine-tune, eval and batch-api inputs, plus job outputs) with size, type, purpose and validation status. Together: GET /files.
Annotations the server declares
- title: List files
- readOnlyHint: true
together_list_fine_tune_eventsList the event log of one fine-tuning job (queued, started, checkpoint saved, epoch completed, errors).No findings
Why it is in this group
- The name "together_list_fine_tune_events" says "list": a word for reading.
Description, as the server wrote it
List the event log of one fine-tuning job (queued, started, checkpoint saved, epoch completed, errors). The first place to look when a job failed. Together: GET /fine-tunes/{id}/events.
Annotations the server declares
- title: List a fine-tuning job's events
- readOnlyHint: true
together_list_fine_tunesList fine-tuning jobs with status, base model, output model name and training settings.No findings
Why it is in this group
- The name "together_list_fine_tunes" says "list": a word for reading.
Description, as the server wrote it
List fine-tuning jobs with status, base model, output model name and training settings. Together: GET /fine-tunes.
Annotations the server declares
- title: List fine-tuning jobs
- readOnlyHint: true
together_list_hardwareList hardware configurations for dedicated endpoints with GPU type/count/memory and price in cents per minute.No findings
Why it is in this group
- The name "together_list_hardware" says "list": a word for reading.
Description, as the server wrote it
List hardware configurations for dedicated endpoints with GPU type/count/memory and price in cents per minute. Pass a model to get only compatible configurations with live availability. Together: GET /hardware.
Annotations the server declares
- title: List hardware
- readOnlyHint: true
together_list_modelsList Together's models with type (chat, language, code, image, embedding, moderation, rerank), context length, organization, license and per-token pricing.No findings
Why it is in this group
- The name "together_list_models" says "list": a word for reading.
Description, as the server wrote it
List Together's models with type (chat, language, code, image, embedding, moderation, rerank), context length, organization, license and per-token pricing. Together: GET /models.
Annotations the server declares
- title: List models
- readOnlyHint: true
together_whoamiIdentify the API key: its organization, project and project slug.No findings
Why it is in this group
- The name "together_whoami" says "whoami": a word for reading.
Description, as the server wrote it
Identify the API key: its organization, project and project slug. The project slug forms the `<project_slug>/<endpoint_slug>` model name for dedicated-endpoint inference. A cheap way to confirm the key works. Together: GET /whoami.
Annotations the server declares
- title: Who am I
- readOnlyHint: true

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Install

Pinned to version 1.16.3. Values in angle brackets are yours to fill in.

claude mcp add --transport http together-ai https://together-ai.usefulapi.io/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json

{
  "permissions": {
    "allow": [
      "mcp__together-ai__together_whoami",
      "mcp__together-ai__together_list_models",
      "mcp__together-ai__together_list_files",
      "mcp__together-ai__together_get_file",
      "mcp__together-ai__together_list_fine_tunes",
      "mcp__together-ai__together_get_fine_tune",
      "mcp__together-ai__together_list_fine_tune_events",
      "mcp__together-ai__together_list_batches",
      "mcp__together-ai__together_get_batch",
      "mcp__together-ai__together_list_endpoints",
      "mcp__together-ai__together_get_endpoint",
      "mcp__together-ai__together_list_hardware",
      "mcp__together-ai__together_list_evaluations"
    ],
    "ask": [],
    "deny": [
      "mcp__together-ai__together_chat_completion",
      "mcp__together-ai__together_create_embeddings",
      "mcp__together-ai__together_generate_image",
      "mcp__together-ai__together_create_fine_tune",
      "mcp__together-ai__together_cancel_fine_tune",
      "mcp__together-ai__together_create_batch",
      "mcp__together-ai__together_cancel_batch",
      "mcp__together-ai__together_create_endpoint",
      "mcp__together-ai__together_start_endpoint",
      "mcp__together-ai__together_stop_endpoint"
    ]
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. It assumes the server is installed as together-ai. Review it before you use it.

Keep track

Get told when this server's tools change: a new tool, a new description, a new class.

https://anlyon.dev/servers/io.usefulapi/together-ai/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Anlyon alternative

10 of these tools change something. Run those as governed actions.

Declare what an action changes, and Anlyon puts it under a shared limit, binds every approval to the exact request, and grades the outcome. It works on any HTTPS API you can describe as an action.

Anlyon does not run, proxy or govern this MCP server. A tool your own code calls directly is not routed through Anlyon and is not governed by it.

Import an OpenAPI spec →

Compare with other servers for the same product.

Source: MCP Registry · repository. Synced Oct 21, 2026, 19:23 UTC.

For agents: Markdown · JSON · report this listing