Blog · AI Agent News · Agent basics VIEN

Gemini 4 Neon: Medium Costs 8× Less Than Claude, Scores Higher

Zapier ran AI through 657 simulated office tasks: Neon on Medium cost $0.58 a task, Claude Fable 5.2 cost $4.64 and scored lower. Which setting should you pick?

Gemini 4 Neon: Medium Costs 8× Less Than Claude, Scores Higher
Images: Photographer: Lukasz Kobus (European Commission) / Wikimedia Commons (CC BY 4.0)
In this article
  1. Does paying 8× more buy a higher score?
  2. Going from Medium to XHigh costs nearly 3× more. What do you get?
  3. What to do about it

Say you have an agent that runs every night to sort your email and sync your spreadsheets (an agent is AI that does many steps for you, like an assistant who searches, compares prices, then books the ticket). I also tend to switch on the priciest model at its deepest thinking level, just to feel safe. Yet in Zapier's test, Claude Fable 5.2 cost about $4.64 a task and still scored below Gemini 4 Neon's $0.58 setting (51.5% vs 53.3%).

My take: pricier does not mean better. Start at Medium, and move up only when it gets things wrong. "Reasoning level" means how long the AI is allowed to think before it answers. The longer it thinks, the more it costs, like a phone plan billed per second.

One limit up front: this is a simulated test (like an AnTuTu score when comparing phones) that Zapier ran on November 17–18, 2026, not your real work. The prices are also estimates from list prices, not invoices.

Neon on Medium costs $0.58 a task, Claude Fable 5.2 costs $4.64
In Zapier's simulated test, Neon on Medium passed 53.3% at $0.58 a task, while Claude Fable 5.2 passed about 51.5% at $4.64 (estimated prices, Nov 2026).Image: Norman P. Jouppi, George Kurian, Sheng Li, Peter Ma, Rahul N / Wikimedia Commons (CC BY 4.0); Derrick Coetzee from Berkeley, CA, USA / Wikimedia Commons (CC0)

Does paying 8× more buy a higher score?

No, at least not in this test. Zapier gave each setting 657 tasks in a simulated company and graded the final result. Neon on Medium got 350 tasks right, while Claude Fable 5.2 at its top setting got fewer right and cost 8× more.

ChoiceScoreCost per task
Neon, Medium53.3%$0.58
Neon, XHigh56.3%$1.69
Claude Fable 5.2 (Max, with Anthropic's default fallback model)51.5%$4.64

Going from Medium to XHigh costs nearly 3× more. What do you get?

Neon on XHigh scores 56.3%, the highest on the board. But measured from Medium, you pay nearly 3× more ($0.58 to $1.69) for about 20 extra tasks out of 657, roughly 3 points. At this level a failed task also costs more than a passed one, so raising the setting without measuring is an easy way to waste money.

Bar chart of pass rates on 657 simulated tasks, with the estimated cost per task of each choice
Pass rate on 657 simulated tasks in Zapier's test (Nov 2026), with estimated cost per task. Neon on Medium is highlighted.

In some places raising the setting is nearly pointless: on operations tasks, Neon's lowest setting already matches its highest. And HR tasks are Neon's weakest spot, where a higher setting does not rescue it.

Also readApple Music Eats 8–11 GB With Zero Downloads: How to Fix It

What to do about it

  1. Open your agent's settings and find which steps run at the highest level.
  2. Move the repetitive steps (sorting email, syncing spreadsheets) down to Medium, run them for a few nights, and compare against before.
  3. Raise only the steps that really keep failing to High or XHigh.

Google also says Neon on High costs roughly 50–80% less per task than other top models, but that is Google's own figure, measured on other leaderboards, not Zapier's.

The comfort of switching on the top setting is not cheap, and by this test it does not even buy a higher score. Next time before you switch it on, ask: does this step really need to think that long?

Sources

  1. Gemini 4 Neon AutomationBench Results: 56.3% Zapier
  2. AutomationBench: AI Agent Benchmarks Zapier
  3. Introducing Gemini 4 Neon Google

Agent basicsGoogle GeminiAutomation

Was this useful?

More on this topic