AI models

Gemini vs Claude

Gemini and Claude API models at the top, mid and budget tiers, with benchmark scores and official prices per 1M tokens.

Gemini 4 Carbon

US$ 2.00 / US$ 10.00Current price, valid through November 30, 2026 Cache read US$ 0.20 Batch US$ 2.00 input and US$ 10.00 output submitted data not used for training

Input US$/1M / Output US$/1M · Official pricing

Claude Fable 5.2

US$ 8.00 / US$ 40.00 1 million token context at the same price · US only ×1.1

Input US$/1M / Output US$/1M · Official pricing

GPQA Diamond

96.5% · 93.9%

Gemini 4 Carbon · Claude Fable 5.2

01

Direct answer

Direct answer: GPQA Diamond: Gemini 4 Carbon 96.5%, Claude Fable 5.2 93.9% (Epoch AI). FrontierMath (tiers 1 to 3): Gemini 4 Carbon 90.5%, Claude Fable 5.2 95.1% (Epoch AI). SimpleQA Verified: Gemini 4 Carbon 77.7%, Claude Fable 5.2 94% (Epoch AI).
see more

02

Top tier: Pick Claude if scores come first: it leads on FrontierMath and SimpleQA Verified and is in a statistical tie with Gemini 4 Carbon on GPQA Diamond. Pick Gemini if cost matters more: at official prices it charges US$ 2.00 versus US$ 8.00 per 1M input tokens and US$ 10.00 versus US$ 40.00 per 1M output.

Mid tier: Pick Gemini if scores come first: it leads on SimpleQA Verified and is in a statistical tie with Claude Sonnet 5.5 on GPQA Diamond and τ-bench Verified. Pick Claude if FrontierMath matter more, where it leads.

Budget tier: Pick Gemini if scores come first: it leads on GPQA Diamond, SimpleQA Verified and τ-bench Verified. Pick Claude if FrontierMath matter more, where it leads, or if cost does: at official prices it charges US$ 0.10 versus US$ 0.75 per 1M input tokens and US$ 0.50 versus US$ 3.75 per 1M output.

Compare with

How to read

  1. Each row is a Gemini tier: top, mid, economy and budget, from most to least expensive. «Open» sits apart and describes the weights, not the price.
  2. «Current model» is the one listed on Google's official pricing page.
  3. When the current model has no published score yet, the tier's scores come from the «reference model for scores», the newest version in the line with a score.
  4. Prices are in US$ per 1M tokens and carry the source label: «Official lab price», «Third-party provider» or «No current price».
  5. Every number comes with its source and date. «No data» means no measurement exists in the open sources this section uses.

Sources

  • Epoch AI, CC BY 4.0, latest version we checked.
  • OpenRouter Benchmarks, CC BY 4.0, latest version we checked.
  • OpenRouter, dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.

Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.

How to cite

If you use this data, cite the page and the date you accessed it.

Adriano Meirinho. "Gemini vs Claude". meirinho.com.br. https://meirinho.com.br/en/ia/modelos/comparar/claude-vs-gemini/. Accessed November 21, 2026.

Third-party data follows the license of each source. The full table is available as CSV and JSON.

Markdown