Loading changelog
Loading changelog
Product updates
Features, fixes, and infrastructure improvements — published as we ship.
Improvement
Caedral platform is tagged v1.0.1. Official SDKs and the n8n node remain at v1.0.0 — no client or model ID changes are required.
See https://caedral.com/changelog and https://caedral.com/pricing
Improvement
Dashboard billing now shows the active top-up bonus table and previews how much prepaid credit you receive before checkout.
Manage balance: https://caedral.com/dashboard/billing
Feature
Self-serve top-ups and auto-recharge now earn bonus prepaid credits through 28 September 2026. You pay the listed amount; bonus credit is added to balance.
Full table: https://caedral.com/pricing
Improvement
Earlier July changelog rows listed prior chat rates, weekly pool language, and subscription-era Base limits. Those no longer describe production.
Current source of truth:
If you bookmarked older rate cards, refresh from /pricing.
Improvement
caedral.com/models now plots Artificial Analysis–style intelligence, coding, and agentic indices for Caedral Base, Titan, Olympus, and Primordial beside leading public models.
Pick a default with evidence — not adjectives.
Feature
Platform and model health are visible without an API key:
GET /v1/status and GET /v1/status/modelsChecks cover Caedral products and core platform services. Use them in runbooks and agency client reporting.
Feature
Official clients are published at v1.0.0:
npm install caedralpip install caedralgo get github.com/trycaedral/caedral-gon8n-nodes-caedral (Community Nodes)All speak the OpenAI-compatible /v1 surface with caedral-* model IDs.
Docs: caedral.com/docs
Feature
To help agencies ship RAG workflows, caedral-embed and caedral-rerank are free on a limited basis through 28 September 2026.
caedral-embed and caedral-rerankEndpoints unchanged: POST /v1/embeddings, POST /v1/rerank.
Improvement
Specialized creative rates are updated on the same prepaid key:
Same endpoints, same model IDs (caedral-vision, caedral-voice). No SDK changes required.
Improvement
Chat prepaid rates are updated across Titan, Olympus, and Primordial. Cached input is now billed at a lower rate when prompt-cache hits are reported.
Charges debit prepaid balance per request. No subscription pools on the API.
Full table: caedral.com/pricing
Feature
Caedral v1.0 is the production line for agencies that need one AI backend without open-ended vendor invoices.
Protocol stays drop-in familiar. Billing stays hard-capped.
Improvement
We've updated Caedral Base with a more generous structure:
We've also strengthened security across the platform, including content safety checks and additional abuse protections.
Improvement
Caedral Embed and Caedral Rerank now run entirely on Caedral's own inference infrastructure — no third-party model dependency for these endpoints.
As a result, pricing is significantly lower:
Same API surface (POST /v1/embeddings, POST /v1/rerank, model IDs caedral-embed and caedral-rerank). No SDK changes required.
Community links are also updated across the site: Discord (primary support), X, and [email protected].
Improvement
We've refreshed our model lineup and introduced dynamic pricing protection:
Chat tiers
Specialized models
New: dynamic pricing protection All usage-based pricing now includes built-in protection — you're never charged more than our advertised rate, even if underlying infrastructure costs fluctuate.
Weekly pool limits now reset every Saturday at 7:00 AM, giving everyone a predictable, consistent cycle.
Feature
Caedral is now live with:
Free tier included, no credit card required to start.