Skip to main content
AIDiveForge AIDiveForge

Novelcrafter vs Riley

Bottom line

Novelcrafter is paid, while Riley has a free plan plus paid upgrades.

Which one should you try?

Novelcrafter

Paid

Best for Fiction writers needing structured outlining tools

Visit Novelcrafter See full review

Riley

Freemium

Best for Creators and founders

Visit Riley See full review

Pricing, capabilities and platforms side by side. Prices are as of the "Pricing checked" date in the table.

Novelcrafter

Novelcrafter

The platform combines a wiki-style Codex with planning modes and an inline editor, so the same tool that holds your worldbuilding also surfaces it while you write. AI assists are one-shot — scene summarization, character extraction, phrasing suggestions — rather than agents running loose, which means you stay in control of every sentence. The Codex links automatically across books in a series, so a character introduced in book one doesn't have to be manually re-entered for book three. The ceiling appears when your workflow demands API-level access or integration into an external pipeline — there is no API and no self-hosted option, so teams that need to connect Novelcrafter to proprietary tooling hit a hard wall. At that point, teams building bespoke author toolchains move to direct LLM integrations or custom document pipelines instead.

Riley

Riley

Riley uses a @context-pill system: you tag a prompt with pre-built context modules — CEO writing style, company background, target audience — and the model generates copy already shaped to those parameters. You highlight output you want changed, leave inline feedback, and Riley adjusts its next generation accordingly. The workflow is deliberate and bounded: this is a prompt-and-refine chat, not an agent running tasks on its own. The ceiling appears early for teams that need volume — the free tier caps at three pieces, and even paid tiers enforce a monthly piece count. Small creator workflows fit cleanly; content teams publishing at scale will hit the limits and start doing math.

AttributeNovelcrafterRiley
PricingPaidFreemium
Price$4.25/mo$14.99/mo
Pricing checkedJul 1, 2026Oct 19, 2026
Free trial21 daysNot stated
Open sourceNoNo
Has APINoNo
Self-hosted optionNoNo
PlatformsWebWeb
Pros
  • The Codex automatically tracks and links characters, locations, and lore as you write, so you avoid the manual cross-referencing that causes continuity errors across long manuscripts or series.
  • A single Codex can be shared across all books in a series, which means worldbuilding work done in book one carries forward without duplication into book three and beyond.
  • AI assists are one-shot and author-directed rather than autonomous, so you retain full editorial control — no agent is rewriting scenes without a direct request.
  • The vendor describes support for external model connections via services like OpenRouter, meaning you can route AI requests through your own API key and avoid being locked to a single provider when costs or model quality shift.
  • Planning modes are designed to surface plot holes and timeline inconsistencies before they reach a final draft, so structural problems get caught at the outline stage rather than in a manuscript review.
  • @context-pills let you encode voice, audience, and style once and attach them to every prompt, so you stop re-explaining your brand on each generation and drafts arrive pre-shaped to your tone.
  • Inline highlight-and-comment refinement keeps iteration inside the same session, which means you avoid the copy-paste loop of prompting, exporting, editing externally, and re-prompting.
  • Multi-model access on paid tiers gives you the option to route different content types to different underlying models, so you are not locked to a single generation quality level.
  • Context pills support stacking — CEO style, company background, and target audience active simultaneously — so a single short prompt carries the full context load without verbose instructions.
  • Social media import lets you pull your existing posts as training material for context modules, which means your defined style is grounded in what you actually published, not a description of how you write.
Cons
  • There is no API — teams that need Novelcrafter to feed data into an external publishing pipeline, content management system, or custom toolchain hit an absolute wall. The workaround is manual export, which breaks any automated workflow.
  • No self-hosted option exists, so authors or teams with data residency requirements, contractual restrictions on cloud-hosted manuscript data, or air-gapped environments cannot use the platform at all — and there is no roadmap item on the public page suggesting this changes.
  • AI assistance is scoped to one-shot writing tasks inside the editor; there is no facility for agents that run multi-step research, draft-and-revise loops, or autonomous outlining passes. Teams whose workflow depends on that kind of iterative AI collaboration move to direct LLM API setups or platforms built around agent frameworks.
  • The platform is paid-only — there is no permanent free tier — so authors evaluating over a longer horizon than the 21-day trial period must commit to a subscription before they have tested the tool against a full project cycle.
  • The free tier hard-caps output at three pieces with no rollover, so teams evaluating whether Riley fits a real workflow exhaust the trial before testing edge cases — and hit a paywall before the tool has proven itself.
  • Paid tiers enforce a monthly piece count ceiling; content teams publishing daily across multiple channels will reach that ceiling mid-month and either pause production or absorb upgrade costs, at which point teams compare Riley against tools that charge per token rather than per piece and often switch.
  • There is no API and no self-hosted option, which means Riley cannot be embedded into an existing CMS workflow, a CI/CD content pipeline, or any internal tooling — teams that need generation to live inside their own stack have no workaround and will evaluate a different provider.
  • Context pill learning is session-scoped, not persistent model fine-tuning across sessions — if your style guide evolves or you want the model to internalize long-term feedback, you are manually updating the pill rather than the tool learning from accumulated output.

Made up your mind?

Novelcrafter

Paid

Best for Fiction writers needing structured outlining tools

Visit Novelcrafter See full review

Riley

Freemium

Best for Creators and founders

Visit Riley See full review

Related comparisons

Free trials & curated lists

Specs come from each tool's own site and docs (how we test). AIDiveForge is editorially independent.

Local AI tools we tested

Browse tools · Affiliate Disclosure · Privacy · Cookies · · Email Policy · Terms · DMCA