fix(benchmarks): correct Step 3.5 Flash cost-multiplier claim to up to 30x

[TechNickAI]

Addresses codex-connector review on PR #16: based on model-benchmarks/data/model-data.json, Step 3.5 Flash (EQ 69.25, $0.15/M) beats Gemini 3.1 Pro (EQ 68.95, $4.50/M = 30x). Claude Sonnet at 40x ($6/M) has higher EQ (71.7), so claiming '10-40x more' overstates the range. Corrected to 'up to 30x more'.

[claude]

Code Review: fix(benchmarks): correct Step 3.5 Flash cost-multiplier claim

Overview

This is a minimal, focused single-line correction in model-benchmarks/index.html that updates the cost-multiplier claim for Step 3.5 Flash from "10-40x more" to "up to 30x more". The change is grounded in the actual pricing data in model-benchmarks/data/model-data.json.

---

Code Quality & Style

• Correct and precise: The fix accurately derives the multiplier from real data — Step 3.5 Flash at $0.15/M vs Gemini 3.1 Pro at $4.50/M gives exactly 30x, making "up to 30x" the correct upper bound.
• Conservative phrasing: Changing from a range ("10-40x") to "up to 30x" is a better journalistic choice — it doesn't overstate the benefit and is defensible from the data.
• No collateral changes: The fix touches only what it needs to. Nothing else was changed, which is ideal for a targeted correction.

---

Potential Issues / Risks

• Minor: The phrase "up to 30x more" is technically accurate but slightly conservative — the PR description itself notes Claude Sonnet is 40x more expensive with higher EQ. If the intent is to capture the full range vs any more expensive model, "up to 40x" would also be defensible. However, the PR rationale (comparing Step 3.5 Flash directly against Gemini 3.1 Pro on the EQ leaderboard) makes 30x the right anchor for that specific comparison. No change needed — just worth noting the interpretive choice.
• Data drift: The claim is tied to pricing that can change. Consider adding a comment in the HTML or a note in the data file indicating when pricing was verified, so future maintainers know when to re-check.

---

Security & Performance

No concerns — this is a static text-only change with zero security or performance implications.

---

Test Coverage

None needed for a copy/text correction on a static site.

---

Summary

✅ Approve — this is a correct, well-reasoned, minimal fix. The math checks out, the phrasing is honest, and the change is appropriately scoped. Ready to merge.

[TechNickAI]

👎 on the data-freshness comment suggestion: the data file is version-controlled — git history is the audit trail for when pricing was verified. Adding inline HTML comments for data freshness isn't the pattern here. The 40x note is acknowledged; 30x is the right anchor for the Gemini 3.1 Pro comparison per PR description.