SaaS tool / MCP router for agents
Connect agents to many third-party apps (Gmail, GitHub, CRMs) with managed OAuth/MCP.
At a glance — Oct 2026
Updated 2026-10-27 · forward this page: /jtbds/agent-tool-router#briefScores as of 2026-10-19 (8 days old).
- Depot Pick
- Composio · 7.62/10 for this job (curated estimate, high confidence)
DX + capability of tool catalog; watch lock-in.
- Runner-up
- Browserbase · 8.30/10
Scores higher than the pick on: Capability fit, Latency / reliability, Developer & agent DX, Privacy / data retention, Lock-in risk (higher = worse), Maturity / track record.
- When not to use the pick
- Only need browsers or one API — call providers directly.
- Lock-in / migration
- Runner-up Browserbase has lower stored lock-in risk (5 vs 5.8, 0–10, higher = more lock-in). Self-host fallback listed for this job: Composio SDK.
- Sourced price (no comparable unit)
Pick: no sourced price in this job's unit (no_matching_sourced_price).
No cheapest SKU (no_comparable_unit_for_job). No price was invented.
- Free option
- Composio SDK (oss) — source, as of 2026-10-25
- Failure modes to watch
- Identity confusion: Bind every tool call to a user or tenant id. Refuse the call if that id changed mid-task. For a public bot, the operator owns the words the bot shows.
- Hallucinated tool arguments: Reject a call that fails the current schema or names an unknown id, and do that before the tool runs a side effect.
- Instruction drift: Restate the stop condition in the trace. End the run when it is met. Review long traces for steps that do not serve the original task.
- Inter-agent misalignment: Log the assigned subtask next to the worker's stated result. The lead rejects a result that does not match the assignment.
- Observability gap: Alert on a drop in trace ingestion and on certificate expiry for the trace host. Do not treat a missing trace as a successful quiet run.
- Partial tool success: Require a per-item status and a failure count. Success for the task requires the failed count to be zero or explicitly handed to a retry queue.
Cite: Agent Provider Depot, Depot Pick for agent-tool-router: composio, scores as of 2026-10-19. Agents: GET /api/decide?jtbd=agent-tool-router
Free SaaS tool / MCP router for agents options · for orchestrator agents
Composio
DX + capability of tool catalog; watch lock-in.
Pricing (directional): As of 2026-10-24 (composio.dev/pricing, fetched 2026-10-24; 'New pricing applies to signups on or after August 15, 2026'): Hobby $0 — '100,000 tool calls, 50,000 triggers, and 3 team members included free each month'; 'The Hobby plan is har…
Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.
Depot pick for tool/MCP router with managed OAuth across 1500+ apps. Official rate card is now public — Hobby is real free for light agents; Pro is the agent-scale default. Not a compute fabric; pair with E2B/Browserbase when you need sandboxes/browsers.
Quote for SaaS tool / MCP router for agents
Depot Pick stays Composio. Prices below are sourced SKU rows only (as of 2026-10-27). Input and output are not blended. GET /api/quote?jtbd=agent-tool-router
Nothing was compared. no comparable sourced unit for this job; chat input $/1M and other units were not borrowed pickIsPriceWinner is null.
- BAA usd_per_tool_call $0.0003 source as of 2026-10-03
- Direct execution outside a Session usd_per_tool_call $0.0001 source as of 2026-10-03
- IP allowlist usd_per_tool_call $0.0001 source as of 2026-10-03
- Proxy execute usd_per_tool_call $0.0002 source as of 2026-10-03
- Sandbox execution usd_per_tool_call $0.0001 source as of 2026-10-03
- Shared connection tool calls usd_per_tool_call $0.0003 source as of 2026-10-03
- Zero data retention usd_per_tool_call $0.0001 source as of 2026-10-03
- Hobby usd_per_month $0 source as of 2026-10-03
- Composio-managed connected accounts usd_per_connection $0.1 source as of 2026-10-24
- Composio-managed tool calls usd_per_tool_call $0.0005 source as of 2026-10-03
- Composio-managed trigger events usd_per_trigger_event $0.005 source as of 2026-10-03
- Pro usd_per_month $29 source as of 2026-10-03
- Sandbox LLM tokens usd_per_million_tokens $3.75 source as of 2026-10-03
- Tool calls own app usd_per_tool_call $0.0003 source as of 2026-10-03
- Trigger events own app usd_per_trigger_event $0.003 source as of 2026-10-03
- Advanced white-labeling usd_per_connection $0.3 source as of 2026-10-03
3 of 7 candidates have no sourced price row. 7 are excluded from the Matching price comparison (none). no comparable sourced unit for this job; chat input $/1M and other units were not borrowed
Runners-up
Browserbase
Depot default for agents that need a real browser. Pair with Stagehand/Browser Use for the reasoning loop.
Anchor Browser
Strong when bot-detection and authenticated computer-use matter. A/B vs Browserbase/Kernel.
Steel.dev
Pick Steel when you need to self-host browsers or hate SaaS lock-in. Browserbase wins on polish.
When NOT to use the top pick
Only need browsers or one API — call providers directly.
Full ranking (score table)
Ranked with this JTBD's weights. Scores are curated estimates unless marked otherwise. Agents: prefer /api/decide then this table for tie-breaks.
| # | Provider | Overall | Fit | $/perf | Latency | DX | As of |
|---|---|---|---|---|---|---|---|
| 1 | Browserbase | 8.3 | 9.2 | 7.5 | 8.8 | 9.4 | 2026-10-19 |
| 2 | Steel.dev | 8.3 | 8.6 | 8.5 | 8.0 | 8.6 | 2026-09-14 |
| 3 | Nango | 8.1 | 8.4 | 8.0 | 8.2 | 8.6 | 2026-09-14 |
| 4 | Smithery | 7.7 | 8.2 | 8.0 | 7.6 | 8.8 | 2026-09-14 |
| 5 | ComposioDepot Pick | 7.6 | 8.2 | 7.5 | 7.8 | 8.8 | 2026-10-19 |
| 6 | Glama | 7.6 | 7.8 | 8.5 | 7.2 | 8.2 | 2026-09-14 |
| 7 | Anchor Browser | 7.2 | 7.8 | 7.2 | 7.2 | 8.0 | 2026-09-14 |
Score methodology for this job
DX + capability of tool catalog; watch lock-in.
capabilityFit: 25%pricePerformance: 18%latencyReliability: 15%developerDx: 15%privacyPosture: 10%lockInRisk: 7%maturity: 10%