
Theo Lindqvist
Technical reviewer — prompts, filters, privacy
Measures the things people usually argue about: refusal rates, context limits, latency, and what a privacy policy actually permits.
Gothenburg · Covering this category since 2023
2023
Testing since
100+
Prompts per tool
Numbers, filters, terms
Remit
About Theo Lindqvist
Theo Lindqvist handles the part of the work that can be counted. Where a review would otherwise say a tool "gets restrictive", Theo runs the same battery of prompts across accounts and reports how many came back refused, at what point in a session the model started hedging, and how long a scene can run before the tool forgets who the characters are.
The background is QA automation, so the habit is to repeat a test until the result stops being anecdotal. A hundred identical requests per tool, spread across different times of day, is the baseline before a latency claim appears in an article. The same discipline applies to context windows: the claim is checked by feeding a story past the stated limit and finding where details start dropping.
The second remit is privacy and payments, which in this category matters more than in most. Theo reads the terms and the privacy policy end to end, checks what shows up on a card statement, tests whether deletion actually deletes, and notes clearly when a policy reserves the right to use your stories as training data.
Areas of focus
- Prompt engineering for long-form fiction
- Content filter and refusal-rate testing
- Context window and memory limits
- Privacy policies and data retention
- Billing descriptors and cancellation flows
Background
- QA automation background
- 100+ measured requests per tool reviewed
- Fact-checks every number published here
- Reads the full terms on every product tested
Articles by Theo Lindqvist
Pricing
Free AI Erotic Story Generator: How Far You Actually Get Without Paying
Six kinds of "free", measured caps per tool, the hidden costs in the terms, and when upgrading is actually the right call.
· 14 min read
Explicitness
AI Erotic Smut: Explicitness, Filters and What "Uncensored" Really Means
The four filter layers, an explicitness scale applied to real tools, the hard limits nobody crosses, and how to avoid repetitive explicit prose.
· 13 min read