Claude Sonnet 4.6 vs Gemini 3 Pro
Direct comparison between anthropic and google for chat models. Specs, prices, strengths, weaknesses — decide on facts.
At a glance
- · Cheaper: Gemini 3 Pro
- · Longer context window: Gemini 3 Pro at 2,000,000 tokens
- · Claude Sonnet 4.6: Best-in-class for code refactoring and multi-step pull requests
- · Gemini 3 Pro: 2M token context — largest window among top-tier models
Spec comparison
| Spec | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| Provider | anthropic | |
| Input per 1M tokens | $3.00 / 1M | $1.25 / 1M |
| Output per 1M tokens | $15.00 / 1M | $5.00 / 1M |
| Context window | 200,000 tok | 2,000,000 tok |
| Max output | 64,000 tok | 65,536 tok |
| Knowledge cutoff | 2025-07 | 2025-11 |
| Vision | ||
| Audio | ||
| Tool calling | ||
| JSON mode | ||
| Structured output | ||
| Reasoning |
Claude Sonnet 4.6 — strengths
- +Best-in-class for code refactoring and multi-step pull requests
- +Very stable tool-use loops — low hallucination on arguments
- +64k output tokens fit whole files in one response
Gemini 3 Pro — strengths
- +2M token context — largest window among top-tier models
- +True multimodal input: video, audio, image, text
- +Very strong on structured output and JSON-schema enforcement
Claude Sonnet 4.6 — weaknesses
- −Knowledge cutoff July 2025 — no recent web knowledge
- −No native JSON mode — schema enforcement via tool-call trick
Gemini 3 Pro — weaknesses
- −Code generation slightly behind Claude Sonnet 4.6 and GPT-5.4
- −At very long contexts, latency becomes significant