The Short Answer
ibl.ai is the MagicSchool alternative for districts that want AI on infrastructure they control, with FERPA-protected student data inside the district's VPC, and pricing that doesn't scale with teacher headcount. Same workloads (tutoring, lesson planning, IEP drafting, parent communication, writing feedback), runtime inside the district's environment, any LLM the district chooses, no per-teacher or per-student pricing tax.
Why Districts Are Looking for a MagicSchool Alternative
Three pressures push districts to look for alternatives:
1. The per-teacher subscription doesn't fit district procurement. MagicSchool prices per teacher. A district with 3,000 teachers pays at least $75–90K/month before any students benefit — for a tool most teachers will use a few times per week and a few will use heavily.
2. Student data lives in MagicSchool's cloud. Tutoring sessions, IEP drafts, parent communication, writing-feedback transcripts — all FERPA-scope, all sitting in a third party's cloud under a DPA the district counsel re-papers periodically.
3. Model choice and language coverage are MagicSchool's call. MagicSchool selects which model handles which workload. Districts serving multilingual learners (Spanish, Mandarin, Arabic, Haitian-Creole, Vietnamese) need locally-controlled language support — not a vendor's roadmap.
What ibl.ai Does Differently
The runtime executes inside the district's VPC. Same network as the SIS (PowerSchool, Infinite Campus) and LMS (Canvas, Schoology, Google Classroom). FERPA-protected student data never traverses a third party's cloud.
Model-agnostic. Run Claude (any tier), GPT-5, Gemini, Llama 4 (self-hosted), DeepSeek-R1, or Qwen 3 (multilingual). The district sets the routing policy. Specifically for multilingual districts — Qwen 3 self-hosted handles Spanish / Mandarin / Arabic / Vietnamese natively without per-language vendor licensing.
No per-teacher / per-student pricing. Usage-based or flat-rate platform license + GPU. A 50K-student district running 96K tutoring sessions/month pays for the actual work (~$3–6K/month all-in), not 3,000 teacher seats.
Open source agent library. The 12 K-12 agent configurations (tutoring, lesson planning, assessment, writing feedback, content creation, special-education-aware, student-safety-monitoring, family communication, curriculum alignment, professional-development, research, administration) live in the open-source iblai/claws repository. Districts fork them, customize for local standards, deploy through the district's ibl.ai platform.
What ibl.ai Replaces
Same K-12 workloads as MagicSchool, on the district's own infrastructure:
- Tutoring — Socratic guidance across grade levels and subjects; Spanish + Mandarin + Vietnamese via Qwen 3
- Lesson planning — standards-aligned (NGSS, Common Core, state standards) with differentiation strategies
- Assessment — quiz creation, rubrics, formative assessments aligned to learning objectives
- Writing feedback — grammar + structure + argumentation review for student essays
- IEP drafting — goal tracking, accommodation suggestions, compliance documentation
- Parent communication — newsletter drafting, progress updates, conference summaries
- Curriculum alignment — mapping activities to standards
- Student safety monitoring — concerning-language detection with crisis-resource referrals
For the per-session token math + vendor comparison (Khanmigo, MagicSchool, Curipod, Brisk Teaching), see What AI Tutoring Actually Costs in 2026 (K-12 + Higher Ed).
The Cost Math
A 50,000-student district running tutoring + lesson planning + IEP drafting:
| Approach | Monthly cost | Student-data location |
|---|---|---|
| MagicSchool / Khanmigo (~$4–10/student × 50K) | ~$200,000–500,000 | Vendor cloud |
| ChatGPT Edu (~$25/teacher × 3K) | $75,000 | OpenAI cloud |
| Microsoft 365 Copilot Edu ($30 × 3K) | $90,000 | Microsoft cloud |
| Direct Claude Sonnet API (token-priced) | ~$2,931 | Anthropic cloud |
| ibl.ai self-hosted (Llama 4 / Qwen 3) | ~$3,000–6,000 | Inside the district's VPC |
At district scale, per-student specialty AI is ~70× more expensive than ibl.ai self-hosted; ChatGPT Edu is ~30× more expensive — for the same tutoring sessions delivered.
For the segment-wide cost math: AI Cost Math for K-12 Districts: Per-Seat vs Usage-Based in 2026.
FERPA + COPPA Posture
For K-12, two compliance overlays matter:
FERPA (Family Educational Rights and Privacy Act) — student records can't be disclosed to third parties without consent. Self-hosted means tutoring transcripts, IEP drafts, and parent-communication content stay inside the district's existing FERPA-scope perimeter; no third-party AI vendor in the data path.
COPPA (Children's Online Privacy Protection Act) — protections for under-13 users. Self-hosted means the district controls data collection, retention, and processing — no vendor terms of service to translate into COPPA compliance.
For the broader policy framework: COPPA Compliant AI for Schools + AI and FERPA Compliance: What Higher Ed Needs to Know (the FERPA arguments map to K-12).
Run the Numbers
- AI Cost Math for K-12 Districts — segment cost math
- What AI Tutoring Actually Costs in 2026 (K-12 + Higher Ed) — per-session math + vendor comparison
- Claw Agents K-12: 12 AI Agents for Schools — open-source agent catalog
- Self-Hosted AI vs ChatGPT Enterprise for K-12 — deployment comparison
- Qwen 3 for Education: Multilingual AI Tutoring — multilingual model
- What Does AI Actually Cost in 2026? — cross-segment pricing hub
Why Family-Owned and New York Matters Here
A school district's AI vendor relationship is a multi-year commitment that touches FERPA-protected student records, IEP documentation, and parent communication. ibl.ai is family-owned and operated from New York, NY — a long-term partner with a perpetual platform license and no investor exit pressure. The runtime is open source. Student data stays inside the district's network. The math works at a 2,000-student elementary district or a 200,000-student urban system.
The MagicSchool alternative isn't a competing per-teacher subscription. It's the district owning the platform.