Transparent test process
How We Test ToolNova Tools
This page documents our pre-release checks rather than making a marketing promise. Opening a tool is not enough: it must accept a real input, produce the expected output type, return understandable errors and behave consistently with the processing method disclosed on its page.
September 18, 2026 live technical SEO audit
The current 54-tool catalog and the public supporting pages in both languages were scanned across all 136 sitemap URLs in production. This audit verified page availability and core technical SEO signals rather than the conversion engine of every tool.
All 136 URLs returned HTTP 200. No missing title or canonical tag, missing or multiple H1, unintended noindex directive, or duplicate page title was detected.
- HTTP 200 for 136 / 136 public URLs.
- Title and canonical tags present on every URL.
- Exactly one H1 per page and no unintended noindex.
- No duplicate page titles detected.
September 8, 2026 end-to-end check
After the OCR release, all 40 active tools were opened and exercised in production. Relevant workflows used JPG, PNG, WebP, a real HEIC file, Canon CR2, PDF, DOCX, XLSX, PPTX, MP4, WebM, WAV and MP3 samples. The 102 public URLs covering tool and supporting pages in both languages were also scanned for HTTP status and initial rendering.
The image OCR endpoint was separately tested with combined Turkish and English recognition and an HTTP 200 response was confirmed. Test files are kept small enough to represent the supported feature; performance and maximum-file-size checks are evaluated separately.
- Opening and core-operation checks for 40 active tools.
- HTTP 200 and initial-HTML checks for 102 public URLs.
- Exact hydration comparison between server and first browser markup.
What do we verify in a conversion?
On input, we check that the correct extension is accepted, invalid or empty files are rejected and size limits are explained clearly. On output, the downloaded file should not be empty, should have the expected MIME type and extension, should open in a basic reader and should leave the source file unchanged.
Document checks cover page count and readability; image checks cover pixel dimensions, transparency and file type; media checks cover duration and playback; OCR checks cover recognized text, language and DOCX or TXT output. Conversion does not promise perfect visual equivalence: results depend on the source structure and the limits of the processing engine.
Distinguishing browser and server processing
ToolNova tools do not all use the same architecture. Simple image operations and some media workflows can run in your browser. Workflows that require dedicated engines—such as OCR, Office, selected PDF operations, RAW decoding and AI transcription—send the file to the ToolNova server over HTTPS.
A mismatch between the interface disclosure and the actual processing path is treated as a release blocker. In the current OCR, PDF and Office services that create temporary server workspaces, cleanup runs when a request finishes or fails. Even so, we recommend following your own security policy for confidential or internal files.
Page quality and accessibility checks
Alongside functionality, we inspect page titles and descriptions, canonical URLs, Turkish–English alternates, keyboard-accessible controls and mobile layout. An automated check compares server-generated HTML with the first browser output; a mismatch blocks release to prevent a page from shifting and repainting during hydration.
Automated checks cannot replace real user feedback. To help us reproduce a problem with a particular camera, browser or complex document, report the tool name, source format, approximate file size, browser and the exact error shown.