CoGrader
AI essay grading against your own rubric, with the teacher reviewing every score
About CoGrader
CoGrader grades student essays against the rubric you already use, inside the platform where the work already lives. It connects to Google Classroom, Canvas and Schoology, pulls an assignment and its submissions, scores each one against your criteria and drafts the comments. Handwritten work is handled too, which matters more than it sounds: a great deal of secondary school writing is still produced on paper, and tools that only read typed submissions quietly exclude it.
The design decision that makes this defensible is that nothing reaches a student without the teacher approving it. CoGrader produces scores and feedback as a draft; the teacher reviews, edits and releases. Grading is the heaviest recurring workload in secondary teaching and the one most often cited in attrition surveys, so the pull is obvious, but AI marking is contested ground and some districts prohibit it outright. CoGrader states the strongest evidence of any tool in this batch on its own homepage: more than 100,000 teachers across more than 16,000 US schools, over 2.5 million essays graded, a U.S. Department of Education IES research award, and a field study in a Texas district. It also claims its scoring has been validated against official state test raters. The pricing has a hole in it, and that needs saying up front: the free tier is generous at 150 submissions a month, the Individual plan that raises it to 350 publishes no price at all, and everything above that is a quote.
You connect CoGrader to your learning platform and import an assignment along with the rubric you intend to mark it against, or build the rubric in the tool. Student submissions come across, typed or scanned from handwriting. CoGrader scores each submission criterion by criterion and drafts feedback comments tied to specific places in the text, so a score is accompanied by a reason rather than appearing as a bare number. The teacher then works through the results: adjusting scores, rewriting comments, overriding anything the model read wrongly. Only when you release them do students see anything. Class-level analytics aggregate the results so that a pattern across thirty essays - everybody is losing marks on evidence, nobody is losing marks on structure - is visible without reading for it, which is what turns a marking session into something you can teach from next week.
- •Grades Against Your Own Rubric - Scores and comments produced criterion by criterion from the rubric you already use, not a generic scale
- •LMS Integration - Google Classroom, Canvas and Schoology, so assignments and submissions move without export and re-upload
- •Handwritten Submissions - Scanned paper work is read and graded alongside typed work
- •Teacher Review Before Release - Every score and comment is a draft until the teacher approves it
- •Class Analytics - Aggregate performance across a class by criterion, to show where the teaching needs to go
- •Generous Free Tier - 150 student submissions a month at no cost, with rubrics and integrations included
Secondary and higher education teachers marking written work at volume, and schools looking to cut the grading load without removing the teacher from the loop. The free tier is large enough to be a real working allowance for one teacher rather than a demo. Against it: the paid ladder reads as "free, then call us", since the Individual plan's price is genuinely unpublished; the product is firmly US-centric in its integrations, its standards framing and its research evidence; and AI marking remains a policy question in many districts, so the tool may be a non-starter regardless of how well it works. Not a developer tool.
Pricing
Not published by the vendor
- Starter$0/mo
- IndividualNot listed
- Schools and districtsContact sales
- Higher Ed and EnterpriseContact sales
From the vendor pricing page, 2026-10-14
