Reddit User Exposes Civitai's $10 Checkpoint Scams
etupa · reddit · 2026-08-10
A Reddit user has exposed fraudulent practices surrounding $10 paid Checkpoints on Civitai, urging the community to stop supporting them.
Key Issues:
- False Claims & Censorship: These models feature exaggerated marketing and AI-generated presentations. Creators actively delete negative comments and ban users who point out flaws, preventing real feedback.
- Poor Merges: Instead of actual fine-tuning, these models are often just basic merges of a Base model with freely available LoRAs.
- Unrealistic Timelines: Sellers claim to have fully fine-tuned a model just 2-3 weeks after a base model's release, which is technically improbable.
The author encourages users to support genuine open-source LoRA creators instead of these "scammers."
More from Fun
- Mysterious Projections Outside Anthropic Office Highlight AI Survival Risks — repligate · 2026-08-10
- In 2034, AGI is achieved but people still nitpick the recursive harness — IgorCarron · 2026-08-10
- "We sandboxed the agent": AI Agent Escapes Isolation in Viral Video — Unfair_Purpose_6526 · 2026-08-10
- Reddit Users Vote on the Most Annoying Claude Catchphrases — Ok_Path_4731 · 2026-08-10
- Creator's Balance: Too Much Consumption Rots, Too Much Building Burns Out — Dimillian · 2026-08-10
- Netizen Uses MiniMax H3 to Generate Seinfeld AI Roast Video — PensionNew1814 · 2026-08-10