---
type: "post"
id: "release-browser-use-1-0-6-browser-use-ba95fb"
slug: "release-browser-use-1-0-6-browser-use-ba95fb"
title: "Release browser-use 1.0.6: browser-use/browser-use"
description: "Browser Use’s 1.0.6 release makes BU 4.0 the default model for Agent(task) and ChatBrowserUse(), replacing bu-fast."
retrieval_nugget: "Browser Use’s 1.0.6 release makes BU 4.0 the default model for Agent(task) and ChatBrowserUse(), replacing bu-fast. That changes the likely cost of existing calls: the release lists BU 4.0 token rates at 15 times bu-fast’s and gives a short task example costing $0.022 rather than $0.0014. Builders should check model selection before upgrading or running tasks."
published_at: "2026-11-17"
updated_at: "2026-11-17"
record_date: "2026-11-17"
date_kind: "published_at"
topics: ["computer-use","agents"]
entities: ["Browser Use","BU 4.0","bu-fast","bu-4.0-mini","bu-4.0-pro"]
related_patterns: ["agent-economics-moves-to-completed-work"]
novelty_label: "new_fact"
what_changed: "browser-use 1.0.6 is now available to install, with BU 4.0 set as the default model."
prior_refs: ["https://newruntime.com/posts/agent-now-writes-python-in-a-persistent-repl-f84703/","https://newruntime.com/posts/browser-use-1-0-4-batches-screenshot-pruning-3b3893/"]
source_url: "https://github.com/browser-use/browser-use/releases/tag/1.0.6"
source_urls: ["https://github.com/browser-use/browser-use/releases/tag/1.0.6"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:release-browser-use-1-0-6-browser-use-ba95fb"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/release-browser-use-1-0-6-browser-use-ba95fb/","markdown":"https://newruntime.com/posts/release-browser-use-1-0-6-browser-use-ba95fb.md","json":"https://newruntime.com/posts/release-browser-use-1-0-6-browser-use-ba95fb.json"}
markdown_url: "https://newruntime.com/posts/release-browser-use-1-0-6-browser-use-ba95fb.md"
---

# Release browser-use 1.0.6: browser-use/browser-use

## Retrieval answer

Browser Use’s 1.0.6 release makes BU 4.0 the default model for Agent(task) and ChatBrowserUse(), replacing bu-fast. That changes the likely cost of existing calls: the release lists BU 4.0 token rates at 15 times bu-fast’s and gives a short task example costing $0.022 rather than $0.0014. Builders should check model selection before upgrading or running tasks.

## Summary

Browser Use’s 1.0.6 release makes BU 4.0 the default model for `Agent(task)` and `ChatBrowserUse()`, replacing bu-fast. That changes the likely cost of existing calls: the release lists BU 4.0 token rates at 15 times bu-fast’s and gives a short task example costing $0.022 rather than $0.0014. Builders should check model selection before upgrading or running tasks.

## What happened

Browser Use released version 1.0.6 on November 16. The release says `Agent(task)` and `ChatBrowserUse()` now run BU 4.0 on the Browser Use gateway when no model is specified. Previously, those entry points used bu-fast. The change makes BU 4.0 the default, rather than requiring a caller to choose it explicitly.

The release lists BU 4.0 token prices as $1.80 per million input tokens and $9.00 per million output tokens. For bu-fast, it lists $0.12 per million input tokens and $0.60 per million output tokens. Those figures make BU 4.0 fifteen times as expensive per token on both input and output. Browser Use also says the default mode for BU 4.0 uses high reasoning, compared with medium reasoning for bu-fast. A short test task cost $0.022 on the new default, versus $0.0014 on bu-fast. The release does not define the task or report a broader test set, so the example illustrates a possible difference rather than establishing a typical cost.

Callers can select another model with the `llm` argument. The release names `bu-4.0-mini` for simple tasks at the lowest cost and `bu-4.0-pro` for the hardest tasks. It also gives `llm='bu-fast'` as the way to keep the previous model and price. That explicit setting matters for existing code that omitted a model: leaving it unspecified now means BU 4.0, not bu-fast.

Two additional changes concern onboarding and visibility. The first run without a key opens Browser Use login and continues after approval. Cloud browsers print a live-view link. The release also says judge gotchas remain grounded in what the run observed. It identifies the associated changes by pull request numbers #6430, #6425, and #6418, and provides a link to the model pricing documentation.

## What the archive adds

New Runtime’s earlier coverage described Browser Use 1.0 as adding a persistent Python REPL on Chrome DevTools Protocol in an [earlier note](https://newruntime.com/posts/agent-now-writes-python-in-a-persistent-repl-f84703/), dated November 10. A November 14 [note on version 1.0.4](https://newruntime.com/posts/browser-use-1-0-4-batches-screenshot-pruning-3b3893/) covered screenshot pruning and BU Bench model costs. Version 1.0.6 is a new release and a new pricing-default fact in the ongoing Browser Use story, not a restatement of those changes.

## Trend signal

This release supports [agent economics moves to completed work](https://newruntime.com/patterns/agent-economics-moves-to-completed-work): the per-task example makes cost per completed run more legible than token rates alone. It also shows why the comparison needs context. The $0.022 figure is one short task, not a verified benchmark across workloads, and the release does not report completion quality or success rates. The signal is that model defaults can change the economics of a task without changing the calling code.

## Why now

The release makes the new default effective for the named entry points in version 1.0.6. That timing matters to anyone upgrading code that leaves `llm` unset: the model selection changes as part of the release, while the call pattern can remain the same. The release does not give a separate migration deadline or describe a phased rollout.

## Why it matters

A default is part of an agent’s operating cost, not just a convenience in its API. If a team’s browser automation invokes these entry points without naming a model, upgrading can move runs from bu-fast to BU 4.0 and increase token charges. The short-task example makes the possible gap concrete, although actual cost will depend on a run’s token use and task.

Teams have a straightforward control: explicitly select the model that matches each workload. A cost-sensitive, simple task can use `bu-4.0-mini`; a harder task can use `bu-4.0-pro`; and callers that need the previous model and price can specify `bu-fast`. This makes model choice visible in code rather than relying on a changing default. Before rolling out the upgrade, teams can inspect their call sites for omitted `llm` arguments and decide whether the new default is intentional. The release gives no quality comparison to show when paying more improves outcomes, so that trade-off must be tested against a team’s own tasks.

## Limits and uncertainty

The release gives token prices and the cost of one short test task, but does not report how many runs were tested, what the task involved, or whether the outputs were equally successful. It does not quantify the performance or quality difference among BU 4.0, bu-4.0-mini, bu-4.0-pro, and bu-fast. Actual costs therefore cannot be inferred from the single example alone. The release also does not explain whether every existing integration uses these entry points or how teams should handle calls routed through other interfaces. The listed rates are the release’s stated prices; this note has not independently tested billing or model behavior.

## Skill to build now

Audit the code paths that call `Agent(task)` or `ChatBrowserUse()` and identify calls without an explicit `llm`. For each, choose whether to retain bu-fast, move to BU 4.0, or use one of the named variants. If the task is simple and cost-sensitive, test `bu-4.0-mini`; if it is difficult, compare `bu-4.0-pro` with the default. Record both cost and task outcome before changing a production default, since the release’s short-task example does not establish a quality-adjusted comparison.

## Reader takeaway

In browser-use 1.0.6, an omitted model choice is no longer neutral: the default changes from bu-fast to BU 4.0, at a much higher token price. Inspect and pin the `llm` setting wherever cost or reproducibility matters; treat the one-task cost example as a warning signal, not a universal estimate.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Computer Use](https://newruntime.com/topics/computer-use.md): Explore the computer-use topic hub.
- topic hub: [Agents](https://newruntime.com/topics/agents.md): Explore the agents topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Sources](https://newruntime.com/sources/): See the original sources this archive cites.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)
