Apps

Real-time analytics for your Swarms workspace: agent counts, executions, success rate, and rate-limit windows across minute/hour/day.

API request log explorer backed by /v1/account/logs. Search and paginate every agent and swarm run with token usage and cost breakdowns.

Charts of input/output tokens and cost across your completions, by day, week, or month, grouped by agent or API key, with filters, a table view, and CSV export.

Paste or upload up to 500 tasks, pick an agent configuration, and fan them out through the batch completions endpoint in chunks of 50. Results stream into a table with per-task status, output, tokens, and cost, exportable as CSV.

Send up to 50 tasks and 50 agents through the batched grid workflow endpoint in one call, then read the results as a matrix: tasks are rows, agents are columns. Click any cell for the full output, and export the whole grid as CSV.

Author single agents (model, prompt, temperature, tools) and execute them against the production Swarms API in one click.

Catalog of agent configurations from /v1/agents/list. Search, sort, and inspect every persisted agent on your account.

Interactive composer for multi-agent swarms. Pick agents, pick an architecture, send a task, and inspect the output.

Build a roster of agents and hold an ongoing conversation with it. One agent chats directly with full memory of the conversation; add more and every agent answers each turn with the transcript as context.

Describe a task and a builder agent designs the team for you — complete AgentSpec configurations with names, prompts, and models you can post unmodified to the swarm endpoints. Backed by /v1/auto-agent-builder/completions.

Configure or import 2 to 6 agents, send them the same task, and compare their responses, latency, tokens, and cost side by side.

Drag agents onto a canvas, connect them into a directed flow, then run the workflow on the platform or export the exact /v1/graph-workflow/completions request as Python, TypeScript, Go, or cURL. Pro/Ultra plans.

A private library for your prompts and skills. Write them here or import Anthropic-format SKILL.md files, markdown, or JSON by drag and drop; tag, search, and reuse them in any agent. Everything is encrypted and visible only to you, and each entry gets its own page.

Deploy agents and services from a GitHub repository, a Dockerfile, or a pre-built image. S2A builds the image, picks the cheapest region across AWS, Google Cloud and Azure, gates the rollout on a healthcheck, and serves it behind a stable URL that scales to zero when idle. In beta: deploy from the Deployments page.

Live model catalog from /v1/models/available: GPT, Claude, Gemini, Llama and more. Cached for 10 hours; refresh on demand.

Every architecture from /v1/swarms/available with descriptions: Hierarchical, Sequential, Concurrent, Mixture of Agents, Council, Debate, Router, Auto Builder, and more.

Every reasoning agent on /v1/reasoning-agent/completions: Self-Consistency for majority-vote answers, Reflexion for self-critique and refinement, and Agent Judge for evaluating outputs. Architecture diagrams, use cases and quick starts.

Build with the Swarms API in any language. Official, typed SDKs for Python, TypeScript, Go, Java, and C#, plus the MCP server for Claude Desktop and Cursor.

A hosted Model Context Protocol server for the Swarms API at mcp.swarms.world/mcp. Live endpoint status and observed uptime, connection snippets for Python and TypeScript, swarm completion examples, and client tutorials.

A specialized Prompt Architect agent (claude-sonnet-4-6) turns a one-line brief into a drop-in deployable system prompt with named sections and operating constraints.

Free, Pro and Premium side by side: which endpoints and models each plan reaches, rate limits, compute lane and support. Every plan pays the same $6.50/M in and $18.50/M out on every model, so a subscription buys access and throughput rather than cheaper tokens.

Token-based calculator for unified pricing ($6.50/M in, $18.50/M out), agent cost, image, MCP, Exa search, web scraper, plus night discount and Frenzy Mode.

Per-minute, per-hour, and per-day windows with live usage bars, countdowns to each reset, tier details, and upgrade options.

Manage the API keys that authenticate your apps against the Swarms API. Full keys are shown once at creation; revoke compromised keys instantly.

Manage your Swarms API key, view your credit balance (paid, free, referral), and choose the workspace theme.

Eight worked calls against the Swarms API — single agents, agents with images, concurrent and sequential swarms, hierarchical swarms, batch completions, and account calls — each with a runnable request in cURL, Python, TypeScript, Go, and Rust.

The Swarms Cloud in one page: capabilities, all 16 swarm architectures, the management console, global infrastructure, pricing, and SDKs.