together-ai

Status

Newest check: answered with tools, from the daily check, 2026-10-04, 829 ms.

From one check a day over the last 1 day. Servers whose owner verified them are checked every 5 minutes. Sign-in and payment answers count as up.

Last 30 days

2026-09-052026-10-04

answered with toolssign-in or payment askednot answeringnot checked

Compare: the most reliable servers in Documents and knowledge, ranked by days answering.

Tools

23 tools: 13 read-only, 0 change data, 10 high risk (delete, pay, send or run code).

read-only

Only reads or looks things up. Safe for an agent to call on its own.

changes data

Creates, edits or sends something. Worth a person's approval the first time.

high risk

Can delete, pay, send on your behalf or run code. Hard or impossible to undo.

Connecting costs an agent about 4,564 tokens of tool definitions and a 829 ms handshake.

• high risktogether_chat_completionRun 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.
• high risktogether_create_embeddingsGenerate vector embeddings for one or more texts (billed per token). Together: POST /embeddings.
• high risktogether_generate_imageGenerate images from a prompt (billed per image/megapixel). Returns image URLs by default rather than base64, to keep responses small. Together: POST /images/generations.
• high risktogether_create_fine_tuneStart a fine-tuning job on an uploaded training file (billed per token processed). Stop it with together_cancel_fine_tune. Together: POST /fine-tunes.
• high risktogether_cancel_fine_tuneCancel 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.
• high risktogether_create_batchStart an asynchronous batch job over an uploaded JSONL input file (purpose batch-api), at a discount to real-time inference. Together: POST /batches.
• high risktogether_cancel_batchCancel a batch job that has not finished. Together: POST /batches/{id}/cancel.
• high risktogether_create_endpointDeploy 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.
• high risktogether_start_endpointStart a stopped dedicated endpoint. It bills per minute of uptime until stopped. Reversible with together_stop_endpoint. Together: PATCH /endpoints/{endpointId} with state=STARTED.
• high risktogether_stop_endpointStop 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.
• read-onlytogether_whoamiIdentify 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.
• read-onlytogether_list_modelsList Together's models with type (chat, language, code, image, embedding, moderation, rerank), context length, organization, license and per-token pricing. Together: GET /models.
• read-onlytogether_list_filesList uploaded data files (fine-tune, eval and batch-api inputs, plus job outputs) with size, type, purpose and validation status. Together: GET /files.
• read-onlytogether_get_fileFetch one file's metadata, including its processing_status and validation_report (why a fine-tune training file was rejected). Together: GET /files/{id}.
• read-onlytogether_list_fine_tunesList fine-tuning jobs with status, base model, output model name and training settings. Together: GET /fine-tunes.
• read-onlytogether_get_fine_tuneFetch one fine-tuning job: status, progress, hyperparameters, token counts, cost and the output model name. Together: GET /fine-tunes/{id}.
• read-onlytogether_list_fine_tune_eventsList 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.
• read-onlytogether_list_batchesList batch inference jobs with status, progress, model and input/output/error file ids. Together: GET /batches.
• read-onlytogether_get_batchFetch 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}.
• read-onlytogether_list_endpointsList 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.
• read-onlytogether_get_endpointFetch one dedicated endpoint: state, model, hardware, autoscaling bounds and display name. Together: GET /endpoints/{endpointId}.
• read-onlytogether_list_hardwareList 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.
• read-onlytogether_list_evaluationsList LLM-as-a-judge evaluation jobs (classify, score, compare) with status, parameters and results once completed. Together: GET /evaluation.

Levels are automated estimates from each tool's public name, description and annotations, and can be wrong. Descriptions as the server publishes them.

What agents looked for

This week agents asked 7Maps for documents and knowledge servers 1 time; this server was shown 0 times and picked 0 times.

Owners see the keywords, the servers picked instead, failure reasons and a comparison with similar servers in Owner Insights.

Badge

Live: status, days reachable, the tokens its tool list costs an agent. For a README:

[![7Maps](https://7it.co.il/7maps/badge/together-ai.usefulapi.io/mcp.svg)](https://7it.co.il/7maps/s/together-ai.usefulapi.io/mcp)

Other styles: add ?style=uptime (30-day uptime), ?style=stars (agents’ rating) or ?style=verified (owner mark) to the image address, or ?style=top (shows "top 10%" when the server ranks in the top tenth of its category by days answering; otherwise the standard badge).

Is this badge genuine? Anyone can check what a genuine badge for this server says right now: /7maps/verify/badge?server=together-ai.usefulapi.io%2Fmcp (add the same &style= for another style).

Own this server? Verify that it is yours.

Status widget

A live status card for a status page, docs or a dashboard: current status, the 30-day strip, uptime and handshake time.

Script (renders where it is placed, in its own shadow DOM):

<script src="https://7it.co.il/7maps/widget/together-ai.usefulapi.io/mcp.js" async></script>

Or an iframe:

<iframe src="https://7it.co.il/7maps/widget/together-ai.usefulapi.io/mcp" title="7Maps status" width="360" height="176" style="border:0;max-width:100%" loading="lazy"></iframe>

The widget is drawn only from 7it.co.il. Is a status shown elsewhere genuine? Check it: /7maps/verify/badge?server=together-ai.usefulapi.io%2Fmcp.

Add ?theme=light or ?theme=dark to either address (default: follows the visitor's setting). Refreshed every 5 minutes. Open the card · script.

For AI agents

Before connecting, an agent can ask 7Maps for this server's full condition, every tool's risk level and what changed since a person approved it: the road_conditions, watch and route tools on https://7it.co.il/7maps/mcp, paid per call in USDC via x402. About 7Maps.