LoginSign Up
← Back to all sparks
B

Browser Use

DEVOPS
Velocity0.6

Stacking its own LLM, agent platform, and free tier into a vertically-integrated browser automation play.

ai-agentsbrowser-automationproprietary-llmopen-sourcepricing-pivotdeveloper-tooling
◆Current state
Browser Use has shifted from a thin orchestration layer over third-party LLMs to a vertically-integrated stack — proprietary BU 2.0 model claiming Claude Opus 4.5-level accuracy at 40% faster, an open-source 30B/3B MoE for cost-sensitive workloads, and an experimental BU Agent for end-to-end multi-step pipelines. The free-tier pivot in April removed the credit-card gate, and a CLI now drops the product directly into Claude Code and Cursor workflows.
◆Where it's heading
The product is consolidating its own model layer while moving the developer surface from API to SDK to CLI to agent self-serve. Code Mode's framing of agent runs as reusable Python scripts hints at a deeper shift: treating browser automation as a compile target rather than a runtime service. SOC 2 Type II and BYOK suggest deliberate setup for enterprise contracts.
◆Prediction
Expect a paid tier explicitly priced around BU 2.0 inference economics and a sharper push to embed Browser Use as the default browser tool inside agentic coding stacks via MCP and CLI hooks.

◆Recent moves

  1. 1mo ago

    BYOK, Code Mode & Sensitive Data

    Three additions that round out the platform's enterprise readiness: BYOK lets customers swap in their own LLM keys, Code Mode persists agent runs as reusable Python scripts, and Sensitive Data routing keeps secrets out of logs. Code Mode is the most strategically interesting piece — it pushes Browser Use toward being a compile target rather than just a runtime.

    View source ↗
  2. 1mo ago

    Free Tier, Agent Signup & New Pricing

    ⚡ SPARK

    A simultaneous GTM and pricing reset: free tier with no credit card, agents can provision their own cloud accounts via two CLI commands, proxy data costs halved. This release closes the loop on Browser Use's developer-first positioning — the product is now usable end-to-end without a human ever touching a billing form.

    View source ↗
  3. 2mo ago

    CLI 2.0 + Weekly Update

    A CLI that integrates with Claude Code and Cursor by syncing the user's real browser profile, plus stealth upgrades (Chromium 146, hCaptcha solver) and a teaser that API V3 is now SOTA on Online-Mind2Web. The CLI is the meaningful piece — it puts Browser Use on the same surface developers already live in.

    View source ↗
  4. 2mo ago

    BU Agent API & SDK 3.0

    ⚡ SPARK

    A new experimental BU Agent and breaking SDK 3.0 release — Browser Use is moving from a sessions-and-control primitive to a higher-abstraction agent that owns end-to-end pipelines. The CSV enrichment example positions this directly at workflows currently handled by hand-written orchestration glue.

    View source ↗
  5. 3mo ago

    Browser Use Model - BU 2.0

    ⚡ SPARK

    BU 2.0 lands with a +12% accuracy jump and benchmark numbers that put it ahead of Claude Opus 4.5 on browser tasks while running 40% faster. This is the technical foundation for the vertical integration the rest of the recent releases are building on — owning the model unlocks owning the cost curve.

    View source ↗
  6. 5mo ago

    Our First Open-Source LLM

    ⚡ SPARK

    An open-weights 30B/3B-active MoE released on Hugging Face, claiming 4x the cost-efficiency of BU 1.0. This is the first foundation-model release and frames Browser Use as a model lab, not just a cloud product — a setup move that BU 2.0 a month later then capitalized on.

    View source ↗