Managed browser-agent loop
Want NL goal → browser agent with optional raw CDP, not just a session host.
At a glance — Oct 2026
Updated 2026-10-27 · forward this page: /jtbds/browser-agent-loop#briefScores as of 2026-09-14 (43 days old; Depot marks scores stale after 30 days). The pick is still Depot's current editorial pick; it has not been re-reviewed since. Prices below carry their own dates.
- Depot Pick
- Browser Use Cloud · 7.40/10 for this job (curated estimate, high confidence)
Agent-loop capability over raw browser hosting.
- Runner-up
- Browserbase · 8.30/10
Scores higher than the pick on: Capability fit, Latency / reliability, Developer & agent DX, Privacy / data retention, Maturity / track record.
- When not to use the pick
- You already have an agent framework and only need hosted Chromium.
- Lock-in / migration
- The pick has lower stored lock-in risk than runner-up Browserbase (4 vs 5). Pick exit signals: open source (provider tag `oss`). Self-host fallback listed for this job: browser-use, WebArena.
- Sourced price (usd_per_browser_hour)
Pick's lowest sourced SKU: Browsers $0.02 (usd_per_browser_hour) — source, as of 2026-10-03, re-checked 2026-10-27
Cheapest matching SKU: Browser Use Cloud Browsers $0.02 — source, as of 2026-10-03, re-checked 2026-10-27. The pick is the cheapest.
Price history: first seen 2026-10-03 · last changed no change recorded · history JSON
- Failure modes to watch
- Benchmark protocol mismatch: Store assisted-or-not, subset, and harness version beside every cited score. Refuse a ranking that mixes those fields.
- Collateral side effects: Snapshot the workspace or app state, then diff it against an allowlist. A passing check plus an unexpected diff is a failure.
- Confirmation skip: Gate those tool names on an approval id recorded before the call. A missing approval is a failure even if the downstream tool returns success.
- GUI grounding miss: After each GUI action, compare the resulting screen or accessibility tree to the control the plan named. Stop when they differ.
- 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.
Same-job, same-unit comparisons: Browser Use Cloud vs Browserbase · Browser Use Cloud vs Anchor Browser · Browser Use Cloud vs Steel.dev
Cite: Agent Provider Depot, Depot Pick for browser-agent-loop: browser-use, scores as of 2026-09-14. Agents: GET /api/decide?jtbd=browser-agent-loop
Free Managed browser-agent loop options · All sourced prices, cheapest first · for browser agents
Browser Use Cloud
Agent-loop capability over raw browser hosting.
Pricing (directional): As of 2026-09-14 (browser-use.com/pricing): No subscription. Browsers $0.02/hr (1-min min); residential proxy $5/GB (on by default); direct/BYO proxy $0.20/GB. Agent API = model cost + 20% service fee (or 20% of published rates if BYOK) + b…
Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.
Pick when you want an agent-native browser loop, not just a hosted Chromium. Browserbase still wins pure session polish.
Quote for Managed browser-agent loop
Depot Pick stays Browser Use Cloud. Prices below are sourced SKU rows only (as of 2026-10-27). Input and output are not blended. GET /api/quote?jtbd=browser-agent-loop
Pick is the cheapest matching USD per browser-hour among candidates (0.02).
- Browsers usd_per_browser_hour $0.02 source as of 2026-10-03
- Direct or your own proxy usd_per_gb $0.2 source as of 2026-10-03
- High-quality residential proxies usd_per_gb $5 source as of 2026-10-03
2 of 7 candidates have no sourced price row. 3 are excluded from the USD per browser-hour comparison (browser_hour). min sourced browser-hour among candidates; lowest browser-hour tier per provider; not quality-adjusted; proxy per GB, CAPTCHA solves, monthly plans, and chat tokens were not included
Sourced USD per browser-hour for Managed browser-agent loop
Candidates for this job that already have a matching sourced price (browser_hour), cheapest first. As of 2026-10-27. Each figure is one SKU in that unit, not a blend of input and output, not a cross-modal rate, and not a Depot score. min sourced browser-hour among candidates; lowest browser-hour tier per provider; not quality-adjusted; proxy per GB, CAPTCHA solves, monthly plans, and chat tokens were not included The Depot Pick is among the cheapest matching prices.
| Provider | USD per browser-hour | SKU | Source | As of |
|---|---|---|---|---|
| Browser Use CloudDepot Pick | $0.02usd_per_browser_hour | Browsers | pricing page | 2026-10-03 |
| Steel.dev | $0.08usd_per_browser_hour | Scale browser hours | pricing page | 2026-10-03 |
| Anchor Browser | $0.09usd_per_browser_hour | Browser duration | pricing page | 2026-10-03 |
| Browserbase | $0.1usd_per_browser_hour | Startup browser hour overage | pricing page | 2026-10-03 |
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
You already have an agent framework and only need hosted Chromium.
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 | TestMu AI (ex-LambdaTest) | 7.9 | 8.4 | 7.8 | 8.2 | 8.0 | 2026-09-14 |
| 4 | Sauce Labs | 7.9 | 8.8 | 6.2 | 8.8 | 7.8 | 2026-09-14 |
| 5 | Skyvern | 7.9 | 8.4 | 8.0 | 7.8 | 8.2 | 2026-09-14 |
| 6 | Browser Use CloudDepot Pick | 7.4 | 8.0 | 7.8 | 7.0 | 8.4 | 2026-09-14 |
| 7 | Anchor Browser | 7.2 | 7.8 | 7.2 | 7.2 | 8.0 | 2026-09-14 |
Score methodology for this job
Agent-loop capability over raw browser hosting.
capabilityFit: 25%pricePerformance: 18%latencyReliability: 15%developerDx: 15%privacyPosture: 10%lockInRisk: 7%maturity: 10%