Quick Start
-
Install and activate the plugin
via the WordPress admin or by uploading the ZIP.
-
Navigate to
WooCommerce → AI Review Summarizer.
-
Select a provider (e.g.,
OpenAI, Google, Anthropic) and paste your API
key.
-
Click Save Changes. Available
models will load automatically.
-
Visit any product page - the summary appears
within seconds.
Screenshot coming soon: WooCommerce → AI Review
Summarizer settings page with provider logos, API
key field, and model dropdown.
Installation
Via WordPress Admin
-
Download the plugin ZIP from the
WordPress.org plugin page.
-
Go to
Plugins → Add New → Upload Plugin.
-
Choose the ZIP file and click
Install Now.
- Activate the plugin.
Via FTP / File Manager
- Unzip the plugin archive.
-
Upload the
ai-review-summarizer folder to
/wp-content/plugins/.
-
Go to Plugins and activate
AI Review Summarizer for
WooCommerce.
Requires: WooCommerce 7.0+,
WordPress 6.0+, PHP 7.4+.
Provider Setup
The plugin connects to third-party AI services to
generate summaries. You need an API key from at
least one provider.
Supported Providers
OpenAI
Single-Provider
Key Models: GPT-4.1 Nano, GPT-4.1
Mini, GPT-4o, GPT-4o-mini. Industry
leader with broad multimodal
capabilities.
Google
Single-Provider
Key Models: Gemini 2.5 Flash, Gemini
2.5 Pro, Gemini 2.5 Ultra. Generous
free tier, strong multimodal.
Anthropic
Single-Provider
Key Models: Claude Sonnet 4, Claude
Opus 4, Claude Haiku 3.5.
Best-in-class reasoning, safety &
code generation.
DeepSeek
Single-Provider
Key Models: DeepSeek-V3,
DeepSeek-R1. Open-weight reasoning
models, extremely cost-effective.
Mistral
Single-Provider
Key Models: Mistral Large 2, Mistral
Small 3, Codestral. EU-hosted,
strong multilingual & code
generation.
X (Grok)
Single-Provider
Key Models: Grok-4. Fast, real-time
knowledge via X platform
integration.
Cohere
Single-Provider
Key Models: Command R+, Command R7B.
Enterprise RAG & tool-use
specialist.
Kimi
Single-Provider
Key Models: Kimi K2.5, K2.6.
Moonshot AI — exceptional
long-context handling.
MiniMax
Single-Provider
Key Models: MiniMax-M2.5, M3. Strong
performance in Asian languages &
long context.
Z (GLM)
Single-Provider
Key Models: GLM-5.1, GLM-4.5-Air.
Zhipu AI — strong Chinese/English
bilingual performance.
Groq
Multi-Provider
Key Models: Llama 4 Scout, Llama 3.3
70B, Mixtral. LPU inference —
fastest tokens/sec available.
Perplexity
Multi-Provider
Key Models: Sonar, Sonar Pro, Sonar
Reasoning. Web-aware
search-augmented generation.
OpenRouter
Multi-Provider
Key Models: 300+ models across all
providers. Single API, unified
billing, model routing.
Together AI
Multi-Provider
Key Models: Llama 4, DeepSeek-V3,
Mixtral, Qwen3. Open-source
inference platform with fine-tuning.
HuggingFace
Multi-Provider
Key Models: Llama 4, Mistral, Gemma
3 & thousands more. Serverless
Inference API.
Azure
Multi-Provider
Key Models: GPT-4o, GPT-4.1, models
from Microsoft. Enterprise
compliance, private networking &
RBAC.
Qwen
Multi-Provider
Key Models: Qwen3, Qwen2.5-VL,
Qwen2.5-Coder. Alibaba Cloud —
strong vision & Asian languages.
AWS Bedrock
Multi-Provider
Key Models: Claude, Llama, Titan
& more. Configurable AWS
datacentre regions worldwide.
Ollama
Multi-Provider
Key Models: Llama 3.3, Mistral,
Phi-4, Gemma 3. Run locally — zero
API cost, full data privacy.
How to Get an API Key
Each provider has their own sign-up process.
Generally:
-
Create an account on the provider's website.
-
Navigate to the API keys / developer section.
- Generate a new API key and copy it.
-
Paste the key into the plugin settings and click
Save Changes.
Security note: Your API key is
encrypted with AES‑256‑CBC before being stored in
the database. It is never displayed in plain text
after saving.
Screenshot coming soon: Provider logo grid with
OpenAI selected, API key field with masked dots, and
model dropdown populated.
Model Selection
After entering a valid API key and saving, the
plugin fetches available models from your provider's
API. Models are cached for 24 hours to avoid
unnecessary API calls.
Select the model that best fits your needs:
-
Smaller/faster models (GPT-4.1
Nano, Gemini Flash): Near-instant summaries,
lower cost, great for most stores.
-
Larger/reasoning models (Claude
Sonnet, DeepSeek Pro): More nuanced summaries,
higher cost, better for complex review patterns.
-
Local models (Ollama): Run
entirely on your own hardware - zero API cost,
full privacy.
Custom Instructions
Append your own guidance to the system prompt to
fine-tune how summaries are written. This is
appended after the built-in prompt, so you
can override or extend behaviour.
Examples
-
Be concise (2–3 sentences), neutral tone,
emphasise pros and cons.
-
Write in a friendly, enthusiastic tone for
first-time buyers.
-
Return 3 bullet points listing strengths
and 2 bullet points listing
weaknesses.
-
If reviews vary wildly in sentiment, state
so explicitly.
Prompt Writing Styles
Choose from preset tone-of-voice templates that
shape how the AI writes your summaries:
Professional
Balanced, objective, informative.
Suitable for large e‑commerce
retailers. Avoids marketing language
and emotional wording.
Friendly
Warm, approachable, conversational.
Reads like helpful buying advice
from someone who has read all the
reviews.
Premium Editorial
Polished, sophisticated language.
Feels carefully curated rather than
generated. Suitable for luxury
brands.
Marketing
Engaging style that highlights the
product's strongest recurring
strengths while remaining truthful
to the reviews.
Widget Placement
Choose where the AI summary appears on your product
pages:
| Option |
Hooks Used |
|
After Product Summary
and within the Reviews tab
|
woocommerce_single_product_summary
+ woocommerce_product_tabs
|
| After Product Summary |
woocommerce_single_product_summary
|
| Before 'Add To Cart' button |
woocommerce_before_add_to_cart_button
|
| Within the 'Reviews' tab |
woocommerce_product_tabs
|
| (Hide) |
Disables automatic display entirely.
|
Screenshot coming soon: Product page with AI summary
widget shown above the add-to-cart button, styled
with icon and accent colour.
Icons & Accent Colour
Choose from a curated set of SVG icons and set a
custom accent colour to match your store's branding:
-
Icon Library: Sargam Icons,
Mingcute, Huge Icons, Lucide, Tabler, Phosphor,
Material Design, and more - 12+ icon sets
available.
-
Colour Picker: Full colour
wheel with live preview of the icon in your
chosen colour.
Screenshot coming soon: Icon selection dropdown with
colour picker and live preview showing a purple
sparkle icon.
Theme Style
Toggle between Light and
Dark widget themes so the summary
box integrates seamlessly with any store design.
Light: Semi-transparent
white background with dark text.
Dark: Semi-transparent
black background with light text.
Batch Scheduling
Schedule summary generation for an entire product
category at once:
- Go to the plugin settings page.
-
Select a category from the
Batch Schedule dropdown.
-
Click
Schedule Summary Generation.
The plugin uses Action Scheduler to process products
in staggered batches - 20 products per batch with a
2-minute interval between batches - ensuring your
server and API rate limits are respected.
Note: This will incur API costs
proportional to the number of products in the
category. Review your provider's pricing before
scheduling large batches.
Summary Expiry
Control how long a cached summary remains valid
before regeneration:
-
One Day / One Week / One Month / One
Year
- Periodic refresh.
-
Upon each addition of a new review
- Always up-to-date with the latest feedback.
-
Never - Generate once and
manage manually. Useful if you want full
editorial control.
Security
-
API Key Encryption: All API
keys are encrypted with AES‑256‑CBC before
database storage. Keys are masked in the admin
UI.
-
Settings API: All settings use
WordPress's built-in Settings API with a
sanitize_callback that validates
every field.
-
Nonce Verification: AJAX
requests (batch scheduling) are nonce-protected
and require
manage_options capability.
-
Output Escaping: All front-end
output is escaped with
esc_html(),
esc_attr(), and
wp_kses().
-
No Direct File Access: Entry
points are guarded with
defined('ABSPATH') || exit;.
FAQ
Yes, it is fully compatible with
WordPress multisite installations.
The plugin itself is free. You only pay
your chosen AI provider for API usage.
Most providers offer free tiers that
cover typical store usage. Using Ollama
for local models is completely free.
Summaries are generated asynchronously
via Action Scheduler. The first page
visit queues the generation; a
subsequent reload (within a few seconds)
displays the result. Once generated,
summaries are cached and appear
instantly.
Yes. Simply select a different provider,
enter its API key, and save. Existing
summaries remain cached; new ones will
use the new provider.
No. All API calls happen via Action
Scheduler in the background. Page loads
are not blocked waiting for AI
responses. Cached summaries are served
from the database with zero API latency.
The summary widget will not appear for
products with zero reviews. Nothing is
generated, and no API call is made.
Yes. Keys are encrypted with AES‑256‑CBC
before being stored in
wp_options. They are masked
(••••••••) in the admin UI and never
transmitted to the front-end.
Support
WordPress.org Forum
Ask questions and report issues on
the official support forum.
Visit Forum
GitHub Issues
Report bugs or request features with
full technical detail.
Open Issue
Donate
Support ongoing development and
server costs.
Donate