Sider

Summary Sider is a browser-native multi-model AI workspace from Vidline Inc. built around three current pillars: Chat for understanding content, Claw for automating browser tasks and Code for changing the websites a user visits. It also includes Wisebase, document and media tools, deep/scholar research, image and video generation and cross-device apps. The product has unusually strong direct adoption evidence—its Chrome extension currently shows 4.9/5 from more than 114,000 ratings—yet its credit system is complex and its privacy policy contains unusually long retention terms that deserve close attention. Rating: Coming soon.

Sider's current site describes one AI sidebar for everything a user does in the browser and claims more than 10 million active users.¹ The exact Chrome Web Store listing provides the stronger independent adoption signal: 4.9/5 from about 114,300 ratings, version 5.32.2, updated August 23, 2026.²

The core Chat surface brings GPT, Claude, Gemini and other models beside webpages and lets users summarize, explain and translate web pages, PDFs and YouTube videos.¹³ This is the most defensible product advantage. The underlying models are available elsewhere, but Sider reduces the friction of moving page context into them and switching among providers.

Current model lists change quickly. The August 2026 App Store description includes GPT-5.6 Sol/Terra/Luna, several Claude 5/4.8 variants, Gemini 3.x, Grok 4.5, DeepSeek V4, Kimi K3, GLM-5 and Qwen3.8-Max, plus Sider Fusion for automatic model selection.³ Those names establish current availability, not proprietary Sider model performance.

The current product has shifted beyond passive chat. Sider Claw is presented as a browser agent that can research, compare, extract and organize information across sites the user already uses, while the mobile app describes multi-source workflows that can produce reports, tables, drafts and files.¹³

Sider Code adds a different capability: changing the appearance or behavior of a visited website and saving those modifications for later visits.¹ That makes the current product more like a customizable AI layer over browsing than a conventional chatbot extension.

These agentic features increase the importance of reliability and permissions. The available independent evidence is mostly user-store evidence rather than controlled testing of Claw on repeatable browser tasks, so claims about autonomous success rates should remain conservative.

Sider also includes Wisebase for collecting websites and uploaded files into an AI knowledge base, plus Deep Research, Scholar Research, ChatPDF, slide creation, website generation, image generation/editing, audio-to-text and other tools.⁴³

The breadth can replace several separate extensions. It also makes Sider harder to evaluate as one product because a user doing lightweight page summaries has a very different workload from someone repeatedly running Claw, Deep Research or video generation.

The changelog confirms rapid model and feature turnover. In 2026 Sider added DeepSeek V4 and continued changing model availability and credit costs; late-2025 updates similarly changed credits for Claude and Gemini configurations.⁵ Buyers should expect entitlements and per-task economics to evolve with upstream model costs.

Sider's current pricing page says all plans include all features and that higher plans provide more monthly credits and a more powerful cloud computer.⁴ The current help center says the active plan family is Lite, Pro and Max, with credits deducted according to actual usage and a detailed credit log visible under Billing.⁶

This is a meaningful change from older Starter, Basic, Professional, Unlimited, Plus and Ultra structures. Sider explicitly labels those older plans as legacy in its June 2026 credit documentation.⁶ Historical reviews that complain about an "Unlimited" plan or quote older credit buckets should therefore be treated as evidence about service history, not as a precise description of a new buyer's current plan.

The live web pricing page is dynamically rendered and exposes the plan structure more clearly than the numeric checkout prices to a text reader. Regional mobile pricing is concrete but heterogeneous: the Italian App Store currently lists several in-app purchases, including a €17.99 monthly Pocket plan, €59.99 annual Pocket plan, €9.99 Basic monthly purchase and €22.99 Pro monthly purchase, alongside legacy and higher-priced products.³ Because these labels mix current and legacy mobile SKUs, the live checkout remains the correct source for a user's exact platform and plan.

Sider's June 2026 credit rules say current-plan users are charged based on actual usage and can inspect each deduction in the credit log.⁶ Older fixed tables show why the distinction matters: Deep Research, Scholar Research, website creation, slides and media workflows historically consumed much more expensive credit categories than ordinary chat.

The current pricing page makes the same product-level point without publishing one universal task tariff: every plan includes Chat, Claw, Code and Create, but higher plans buy more credits and cloud compute.⁴

That makes the product attractive to users whose activity is mostly summarization, translation and ordinary chat, but less predictable for heavy browser automation and media creation. The practical buying test is to use the free tier on representative tasks and inspect the actual credit log before purchasing a long annual plan.

Sider can be installed and used free, with paid plans expanding credits and compute.¹⁴ The Chrome extension's very large population indicates that the free/onboarding surface is not merely a private beta or token demo.²

The iOS app is also free with in-app purchases and has 4.7/5 from about 4,300 U.S. ratings; the Italian listing has 4.7/5 from 99 ratings.³ Mobile satisfaction should be treated separately from the much larger Chrome population because the workflows and interface differ.

Chrome Web Store currently reports 4.9/5 from roughly 114,300 ratings for the exact Sider extension.² This is an unusually large exact-product sample and directly supports the claim that many users value the sidebar workflow.

The store also says the extension handles personally identifiable information and website content. The developer declares that data is not sold to third parties, not used for unrelated purposes and not used for creditworthiness decisions.²

The current Apple App Store adds a smaller but still relevant exact-product population at 4.7/5 from about 4,300 U.S. ratings.³

These store scores measure the Sider product experience. They do not establish that a Claude, GPT or Gemini response routed through Sider is more intelligent than the same upstream model used directly.

Sider's current linked privacy policy was last updated March 5, 2025. It says the company collects account information, usage information, device/IP data, communications and content submitted to the service; it may share personal information with service providers including Google, Cloudflare and Microsoft.⁷

The most important disclosure is retention. The policy states that account-related information is retained for 99 years after an account is closed and communication information for 99 years from the last communication, while usage and analytics information is retained for three months before anonymization or deletion.⁷

A 99-year retention period is extraordinary for an AI productivity service and should be a material procurement issue for privacy-sensitive users. The policy also says reasonable safeguards include encryption, access controls and other administrative/technical measures, but does not make the retention concern disappear.⁷

The policy's content-rights section is also unusual: it describes generated content as potentially jointly copyrighted by the user, Sider and third-party service providers while granting the user broad commercial/non-commercial use rights.⁷ Users relying on exclusive ownership of generated assets should obtain legal clarification rather than infer conventional SaaS ownership terms.

Trustpilot is much smaller and far more negative than the product stores. The current sider.ai profile is 2.4/5 from 15 reviews, with 74% one-star reviews.⁸ Recent 2026 complaints focus on customer support, account access, PDF failures, model quality, cancellation and perceived value, while a June 2026 positive review praises the range of models and media generation.

This population is self-selected and unsuitable as a direct product rating. It is useful for a different question: what problems motivate users to seek a public service-review channel? The repeated themes around support, billing and access deserve consideration before an annual or team purchase.

Older Trustpilot complaints about "Unlimited" plans must also be interpreted against Sider's documented 2026 plan migration; they describe a legacy commercial structure, not necessarily the current Lite/Pro/Max entitlements.⁶⁸

Sider is strongest for people who spend much of their workday in a browser and want one persistent AI surface for pages, PDFs, YouTube, email, translation and writing.¹ The extension's scale suggests this is not a niche workflow.

It is also attractive to multi-model users who want rapid access to current GPT, Claude, Gemini, Grok, DeepSeek and other models without keeping several separate chat tabs open.³ Wisebase and research tools make it more useful for people who repeatedly work across collections of documents and web sources.

Claw is potentially valuable for users who want browser automation rather than only answers. Because current independent controlled evidence on Claw reliability is thin, users should start with low-risk, reversible tasks and verify outputs before delegating actions with meaningful consequences.

Sider is less attractive for a user committed to one provider's native ecosystem or someone who needs simple, predictable flat-rate economics. Native ChatGPT, Claude or Gemini products can expose provider-specific features, memory and model controls more directly.

Organizations with strict retention policies should read the privacy policy before deployment. The 99-year account/communication retention language is difficult to reconcile with many data-minimization programs unless separate enterprise terms materially change it.⁷