AI and SaaS Reviews | Future Stack Reviews

Side-by-side price comparison for Hostinger Business hosting. Left card, when you buy: 579 yen per month. Right card, when you renew: 1,899 yen per month, the same rate at 12, 24 or 48 months. A red not-equals sign separates them. Banner beneath reads: the term discount only works when you buy.

Hostinger renewal pricing: the term discount does not come back

July 26, 2026

One paid Hostinger Business account, first term priced against second. The term discount that made the purchase attractive was worth nothing at renewal.

Claude Opus 5 review card reading no scored difference and not drop-in. A diagram shows Opus 4.8 and Opus 5 reaching the same verified result, while the API side is marked unequal with a warning.

Claude Opus 5 vs Opus 4.8: Six Runs, No Detected Difference, and a Migration That Still Needs Work

July 26, 2026July 26, 2026

Six runs, three per model, on one frozen task. Claude Opus 5 and Opus 4.8 matched on every scored field. What that tells buyers, and what it does not.

Gemini 3.6 Flash review summary card contrasting a $7.50 per million output token price cut with migration work marked not a drop-in swap, beside a panel of model specifications and pricing.

Gemini 3.6 Flash Review: The Price Cut Is Real.

July 26, 2026July 23, 2026

Gemini 3.6 Flash review: the output price cut is real, but sampling controls, storage defaults and Batch access change with the API surface you pick.

Claude Science review card: audit trail does not equal accuracy, with a reviewer that checks claims against the record but does not re-run or peer-review.

Claude Science Review: what it is, and how it differs from the regular Claude app

July 21, 2026

Hands-on Claude Science review: same models as the regular Claude app, plus a sandbox, connectors, and provenance. Inspectable by design, correct only where you check.

Editorial graphic contrasting Fable 5 building a landing page with its own zero-fix security review, tagged scope not safety, static file review, and limited by design.

Fable 5 Built a Landing Page, Then Security-Reviewed Its Own Code. Zero Fixes. Here’s What “Clean” Actually Meant.

July 23, 2026July 19, 2026

Fable 5 built a landing page, then security-reviewed its own code for zero fixes. Here is why a clean result was a scope record, not proof the code is safe.

Future Stack Reviews Tier B infographic for Claude Fable 5, July 2, 2026. The same Fable usage is shown three ways with no official conversion between them: 5% weekly in the Claude app, 95.2K tokens in Claude Code, and $0.00 in the usage credits panel. Restored July 1, 2026. API pricing is $10 per million input and $50 output versus Opus 4.8 at $5 and $25, included up to 50% of weekly limits through July 7, then usage credits.

Claude Fable 5 Is Back, But Its Usage Meters Do Not Agree

July 23, 2026July 2, 2026

Anthropic restored Claude Fable 5 on July 1, but it returned as a capped, separately metered entitlement. On a Max 5x account, FSR found tokens in Claude Code, a percentage meter in the Claude app, and usage credits at $0.00, with no official way to convert between them. A Tier B hands-on look at Fable 5’s usage limit, credits, and the cost you cannot yet read.

Instantly AI Pricing Review cover card with verified claim status, $9/150 credits breakdown, and a blocked AI Sales Agent UI.

Instantly AI Pricing Review: What $9 in Credits Actually Buys (and Where It Stops)

June 25, 2026June 17, 2026

Instantly sells Credits and email sending as separate subscriptions. The $9 Nano Credits plan funds AI lead search, enrichment, and agent drafts, but it cannot send a single email on its own. FSR hit that wall: the AI Sales Agent stopped at the sending-account check, not the credit balance.

FSR Tier B review card for Mitte AI, headlined Strong output, a credit meter that is only sometimes right. Across eight billed tests, four of six configurations billed as displayed and two billed higher, the worst case shown 45 and billed 280, with a free-plan overdraft of -180 credits. FSR verdict: worth trying on the free plan first, and budget from receipts, not the panel.

Mitte AI Review: Strong Output, a Credit Meter That Is Only Sometimes Right, and a Price List That Is Not Public

June 29, 2026June 13, 2026

Mitte runs many image, video, and audio models on one credit balance, and the images are strong. The pricing is the part to check before you pay. In our paid testing, two of six configurations billed more than the panel showed on every run, one shown at 45 credits that settled at 280. Budget from the receipt, not the panel.

Zebracat review summary card from Future Stack Reviews, labeled a Tier B buying-layer audit rather than a video-quality score. Its core finding is that the buyer risk sits in the credit meter: refund rules, entitlements, model costs, review incentives, and enterprise paperwork. The card lays out Zebracat's three plan prices, a four-step buyer flow that ends at a policy gap, the one-time credit packs with 60-day expiry, and what FSR tested firsthand versus what it did not score.

Zebracat Review: What Happens If You Buy Credits Before You Subscribe

June 11, 2026June 11, 2026

We bought Zebracat credits before subscribing, and the refund was refused. This Zebracat review audits the buying layer where the risk actually sits: refund rules, credit metering, model costs, review incentives, and enterprise paperwork, all with receipts and dates. It does not score video quality, because we generated none.

Odysseus review cover image: the self-hosted AI workspace interface with the headline "local-first, until you connect a cloud model."

Odysseus Review: Where PewDiePie’s Self-Hosted AI Workspace Actually Sends Data

July 17, 2026June 9, 2026

Odysseus, PewDiePie’s free self-hosted AI workspace, is private only along the paths you keep local. FSR captured the traffic on one machine: the local model sent nothing out, a cloud model left for OpenRouter, and the default setup returned a 404 on first run.