Multi-tenant SaaS, isolated by design
Every project, voice, and audio asset is scoped to a tenant with row-level security. Roles, audit logs, and quotas — nothing leaks between studios.
EPUB → Audiobook
Your library, narrated by your favorite voice.
Crossread turns EPUBs into hand-crafted audiobooks in the cloud — with the TTS provider you pick. Built for authors, archives, and accessibility teams who want their words heard exactly the way they were written.
Built for production
Managed cloudEU-hostedGDPR-nativeEncrypted at restNo install, no opsWhat you get
Everything you need to ship audio.
A studio for narration, a vault for your keys, and an audio engine that streams instead of stalls.
Pluggable TTS providers
ElevenLabs, Google Cloud, OpenAI, Qwen3, VoxLibri. Switch per project, per voice, per section.
Range-streamed audio
HTTP byte-range delivery so listeners scrub instantly — no full-file downloads, no waterfall.
Public share links
Per-project tokens with optional password and expiry. Read-only, audit-logged, revocable.
Encrypted key vault
API keys encrypted at rest with XChaCha20-Poly1305 + per-user HKDF subkeys. Workers decrypt in isolation.
Managed cloud, zero ops
Sign up and start producing — no servers, no Docker, no rotating keys. We host, patch, monitor, and back up the whole stack on EU infrastructure.
The flow
Three steps from manuscript to audio.
Upload your EPUB
Drop the file. Crossread parses spine, chapters, cover, and metadata so you start with clean sections.
Pick a voice + provider
Choose any voice from any supported TTS engine. Override per chapter when the narration calls for it.
Stream, share, archive
Listen with byte-range streaming, share a read-only link, or export the audio for your own pipeline.
Questions, answered
The fine print, plainly stated.
Crossread treats security and tenant isolation as table stakes. If you find a gap, email us — we patch first and write a postmortem second.
How do I get started?
Create a free account in under a minute, confirm your email, and you are inside the studio. There is nothing to install — Crossread is delivered as a managed cloud service, not a download.
Where is my data stored?
Audio, projects, and metadata live on EU-hosted infrastructure. Storage and database are encrypted at rest, traffic is TLS only, and tenants are isolated through Postgres row-level security.
Which TTS providers can I use?
ElevenLabs, Google Cloud TTS, OpenAI, Qwen3 (audio), and VoxLibri are wired in. You pick a provider per project or per chapter — no engine is forced on you and you keep your narration choices.
Is my data tenant-isolated?
Every query runs under Postgres row-level security keyed to the tenant ID. Audit logs, jobs, audio cache, share tokens — every record carries the tenant scope, and our test suite gates merges on it.
What happens to audio when I delete a project?
The cache for that project is evicted within minutes. Hard-delete pipelines purge audio chunks, share tokens, and audit trails in a single transactional sweep that you can verify in the audit log.
Is there a free tier?
Pricing is in flight while we wrap quotas and billing. Reach out — we're onboarding design partners with generous credits and white-glove migration help.