Key Takeaways
- EssayHero supports Cambridge IGCSE English Literature (0475) and English Language (0500), covering three distinct assessment types: Literature essays, Directed Writing, and Composition
- All criteria and level descriptors are drawn directly from official Cambridge Assessment mark schemes, which are publicly available
- The AI cannot verify literary quotation accuracy, cannot assess responses to unseen stimulus material, and cannot replicate the contextual judgement of trained Cambridge examiners
- This is a formative revision tool, not a replacement for teacher assessment or examination marking
If You Teach IGCSE English, Read This First
Cambridge IGCSE is the world's most popular international qualification for 14-16 year olds. If you teach English Literature (0475) or English Language (0500) at an international school, you know the challenge: students need practice with timed essays, they need feedback aligned to Cambridge criteria, and there are never enough hours in the day to provide it at the frequency they need.
EssayHero now supports three Cambridge IGCSE assessment types. This post explains exactly how it works, where the criteria come from, and what it cannot do.
Why Transparency Matters
If you are going to consider recommending it to your students or their parents, you deserve full transparency.
I'm Joseph Lin. I've been marking essays for over twenty years, from primary school through to PhD dissertations. I built EssayHero originally for HKDSE students in Hong Kong.
It expanded to Cambridge IGCSE because teachers at international schools asked for it, and because Cambridge's publicly available mark schemes make alignment straightforward and verifiable. EssayHero is free, open source, and has no commercial aims.
Three Assessment Types, One Page
Cambridge IGCSE English involves different papers with fundamentally different assessment approaches. EssayHero handles three of them:
0475 English Literature
This paper assesses critical analysis of set texts (poetry, prose, drama). Students respond to extract-based or discursive questions.
Assessment structure:
- Response to Text — 15 marks
- Personal Engagement — 5 marks
- Language Analysis — 5 marks
- Total — 25 marks
The criteria map to Cambridge's four Assessment Objectives: AO1 (knowledge of content), AO2 (understanding of meanings and contexts), AO3 (language, structure and form), and AO4 (personal response).
0500 Directed Writing
Section A, Question 1 assesses transactional writing based on stimulus material. It uses a single integrated Writing scale of 25 marks.
The scale covers:
- W1 — Style (conveyance of meaning)
- W2 — Structure (organisation for reader)
- W3 — Vocabulary (range and precision)
- W4 — Register (appropriateness to audience)
- W5 — Technical accuracy (grammar, spelling, punctuation)
This is the paper where students must adapt their writing to a specified purpose and audience.
0500 Composition
Section B assesses creative and discursive writing. It uses two criteria: Content and Structure (16 marks) and Style and Accuracy (24 marks), totalling 40 marks.
Students choose between descriptive and narrative tasks, and Cambridge applies different specific descriptors depending on which form the student selects.
Configuration Philosophy
Each assessment type has its own configuration, its own level descriptors, and its own prompt instructions. They share the same underlying strictness modes (Lenient, Baseline, Harsh) and feedback philosophy, but the actual criteria and mark allocations differ — because that is how Cambridge assesses them.
Where the Criteria Come From
Every criterion, every level descriptor, and every Assessment Objective reference in EssayHero's IGCSE configuration is drawn from official Cambridge Assessment publications. Cambridge mark schemes are publicly available documents, and we use them directly.
Official Sources
For 0500 (Language): Level descriptors are taken verbatim from the published 0500/21 Mark Scheme (May/June 2025).
For 0475 (Literature): Band descriptors follow the generic mark scheme criteria for Literature essay responses published in the 0475 syllabus documentation.
Why This Matters
The feedback your students receive is aligned to the same framework you use when you mark their practice essays. The AI is not applying some proprietary rubric or a generic "good writing" checklist.
It is applying Cambridge's own criteria, at Cambridge's own level boundaries, using Cambridge's own terminology.
Full Verification Available
The complete prompt configuration, including every level descriptor and Assessment Objective reference, is published on our methodology page and in the source code. Nothing is hidden.
What the AI Actually Looks For
0475 Literature
Response to Text (15 marks)
This is the weightiest criterion. The AI assesses whether the student demonstrates detailed knowledge of the text (AO1) and understanding of its meanings, themes, and contexts (AO2).
Level 6 (13-15 marks): The AI looks for "thorough and perceptive knowledge" with "sophisticated understanding" and seamlessly integrated textual evidence.
Level 3 (5-7 marks): It expects some knowledge but notes when interpretation is superficial or inconsistent.
Personal Engagement (5 marks)
This criterion rewards genuine, individual responses. The AI is trained to distinguish between rehearsed, generic observations and original thinking that illuminates the text.
A student who offers their own interpretation, even if imperfect, scores higher than one who reproduces standard critical opinions without engagement.
Language Analysis (5 marks)
This assesses whether the student can move beyond identifying techniques (feature-spotting) to explaining their effects.
At the higher levels, the AI looks for analysis of how writers use language, structure, and form to create meaning, with accurate use of literary terminology.
0500 Directed Writing
The single Writing criterion integrates all five Assessment Objectives holistically. The AI evaluates:
- W1 — Does the style convey meaning effectively?
- W2 — Is the response structured for the reader's benefit?
- W3 — Is vocabulary varied and precise?
- W4 — Is register appropriate to audience and purpose?
- W5 — Are spelling, punctuation, and grammar accurate?
Level 6 expectations: "Highly effective style capable of conveying subtle meaning" with "wide range of sophisticated vocabulary, precisely used."
Level 3 expectations: "Inconsistent style" where "expression [is] sometimes awkward but meaning clear."
0500 Composition
Content and Structure (16 marks)
This criterion assesses ideas and organisation. The AI applies different specific descriptors depending on whether the writing is descriptive or narrative.
Descriptive writing at Level 6: "Many well-defined and developed ideas and images [creating] a convincing overall picture with varieties of focus."
Narrative writing at Level 6: A "well-defined and strongly developed" plot with "features of fiction writing such as description, characterisation and effective climax."
Style and Accuracy (24 marks)
This focuses on the craft of writing:
- Vocabulary precision
- Sentence variety
- Register consistency
- Technical control
Level 6 expectations: "Precise, well-chosen vocabulary and varied sentence structures, chosen for effect" with "spelling, punctuation and grammar almost always accurate."
The Feedback Philosophy
Two principles govern how the AI provides feedback across all three assessment types.
1. Positive Marking
Cambridge's own guidance instructs examiners to "award marks positively for what the candidate demonstrates." The AI follows this principle.
It identifies what a student has achieved before discussing what needs improvement. Suggestions are framed as opportunities to reach the next level, not as deductions from a perfect score.
2. Specificity with Assessment Objectives
The AI references Assessment Objectives (AO1-AO4 for Literature, W1-W5 for Language) when explaining its feedback.
Instead of: "Good vocabulary."
The AI says: "This demonstrates W3 at Level 5: a wide range of vocabulary used with some precision. To reach Level 6, aim for vocabulary that is not only varied but precisely chosen for subtle effect."
This helps students understand what the criteria actually mean in practice.
What We Cannot Do
Critical Limitations
This is the section that matters most. Read it carefully before recommending EssayHero to your students.
We Cannot Verify Literary Quotation Accuracy
For 0475 Literature, the AI assesses how textual evidence is integrated and analysed — its relevance, specificity, and analytical deployment.
But it cannot confirm that a quotation is accurately remembered from the source text. If a student misquotes a passage, the AI will not catch it.
Teacher action required: Check quotation accuracy separately.
We Cannot Assess Response to Unseen Stimulus Material
Directed Writing requires responding to specific passages that the AI does not have access to.
The AI evaluates the quality of writing, register, and structure, but it cannot judge whether the student has accurately addressed the content of the stimulus material.
We Cannot Replicate Cambridge Examiner Judgement
Cambridge examiners undergo rigorous standardisation training for each examination session. They bring collective calibration that an AI, however well-configured, cannot match.
The AI applies the published criteria consistently, but it lacks the contextual judgement that comes from marking thousands of scripts within a specific session.
We Cannot Assess Speaking and Listening
Cambridge IGCSE English includes speaking and listening components that are entirely outside the scope of written essay assessment.
Scores Are Indicative, Not Predictive
When Teacher and AI Disagree
If a student receives Level 5 on EssayHero and Level 3 from their teacher, the teacher is right.
The AI provides a benchmark against the published criteria, but it does not account for:
- Your specific department's marking conventions
- The particular question set
- The standardisation adjustments that apply in a live examination
What It Is Good For
After that list of limitations, the reasonable question is: why bother?
The Answer: Iteration Speed
A student preparing for Cambridge IGCSE examinations needs to practise timed essays, receive feedback, revise, and practise again.
Most students do not get enough of this cycle because teacher feedback, however good, is limited by time.
The Practice Cycle
EssayHero enables this workflow:
- Student writes a Literature essay at home
- Receives paragraph-by-paragraph feedback aligned to Cambridge's AO1-AO4
- Identifies that their language analysis is at Level 3 because they are feature-spotting rather than explaining effects
- Revises that section
- Resubmits
The feedback is consistent — the same essay submitted twice will receive the same assessment — which helps students understand what the criteria actually require.
Strictness Modes for Calibration
The strictness modes are particularly useful for Cambridge students:
Baseline mode: The AI applies standard Cambridge marking.
Harsh mode: Rigorous standards where Level 6 marks are reserved for genuinely outstanding work. A student who consistently scores Level 4-5 on Harsh mode can feel confident going into the examination.
Lenient mode: One who struggles to reach Level 3 on Lenient knows exactly where to focus their revision.
For Parents
EssayHero Is a Practice Partner
EssayHero provides the kind of structured, criteria-referenced feedback that helps a student understand why they received a particular score, not just what score they received.
It is a practice partner, not a tutor.
Full Transparency
Published Criteria and Prompts
The complete assessment criteria, all level descriptors, and the full AI prompt instructions for each Cambridge IGCSE paper are published on our methodology page.
Cambridge's mark schemes are public documents, and we use them directly. You can compare our configuration against the official syllabus documentation and verify the alignment yourself.
Open Source Codebase
EssayHero is open source under AGPL-3.0. The complete codebase, including every prompt section, is available at github.com/smartjolin/essayhero.
Privacy Commitment
Essays are processed and discarded. They are not stored, not used for model training, and not accessible to anyone after feedback is generated.
Privacy is straightforward: we process the essay, generate feedback, and forget it.
Try It Yourself
EssayHero is free. No account required.
Test Each Assessment Type
If you want to see what the feedback looks like on a Cambridge IGCSE essay:
- Literature (0475) — Go to essayhero.app/?exam=caie-igcse-0475, paste a practice essay, and read the output
- Directed Writing — Select "IGCSE 0500 Directed" from the exam selector
- Composition — Select "IGCSE 0500 Composition" from the exam selector
Feedback Welcome
If you think it could help your students prepare more effectively, share it.
If you think the criteria alignment could be improved, or the feedback is not useful for your particular context, I would genuinely like to hear from you. Email hello@essayhero.app.
Why This Exists
I built this to help students write better. Cambridge IGCSE students at international schools around the world deserve access to quality feedback between drafts, regardless of whether their school can provide it at the frequency they need.
If EssayHero can fill that gap, even partially, it has done its job.
EssayHero is free, has no commercial aims, and is built by a Hong Kong teacher for students worldwide. Questions? Email hello@essayhero.app.
Related Articles
How EssayHero Marks Business Essays (And What It Can't Do)
A transparent look at how EssayHero assesses university business and management essays, what the criteria actually measure, and where AI falls short.
Read moreHow EssayHero Marks HKDSE Paper 2 Essays (And Why You Should Know)
A transparent explanation for teachers and tutors of how EssayHero assesses HKDSE English Paper 2 writing, how scoring works, and where AI falls short.
Read moreHow EssayHero Marks IELTS Essays (And What It Can't Do)
A transparent look at how EssayHero assesses IELTS Writing Task 1 and Task 2 essays, what the four criteria actually measure, and where AI falls short of a real IELTS examiner.
Read more