Gemini 3.1 Pro vs Claude Sonnet 4.6: the developer API cost face-off
Gemini 3.1 Pro is more economical on short-context base and cached costs. Claude Sonnet 4.6 is preferred if you need Anthropic's coding and instruction-following strengths or its pricing fits your prompts above Gemini's 200K tier boundary.
Coding workhorses and reasoning model showdown.
For complex coding and agentic reasoning, Gemini 3.1 Pro and Claude Sonnet 4.6 are the top contenders. Their cost structures depend heavily on prompt caching efficiency.
By TechCompare · Updated
Cost Comparison
Based on 100,000 input tokens (50% cached), 5,000 output tokens, and 100 requests.Prices are fetched live from OpenRouter and may include temporary promotional discounts not accounted for in our article and comparison figures.
Side-by-side specs
| Spec | Gemini 3.1 Pro (<=200k) | Claude Sonnet 4.6 |
|---|---|---|
| Input Cost (per M) | $2.00 (better on this spec) | $3.00 |
| Output Cost (per M) | $12.00 (better on this spec) | $15.00 |
| Cached Input (per M) | $0.20 (better on this spec) | $0.30 |
| Batch Discount | 50% | 50% |
How they differ
For prompts up to 200K tokens, Gemini 3.1 Pro costs $2.00 per million input tokens and $12.00 per million output tokens, with a 90% cache discount to $0.20/M. Its rates rise to $4.00/M input, $18.00/M output, and $0.40/M cached input above 200K. Claude Sonnet 4.6 is priced at $3.00 per million input tokens and $15.00 per million output tokens, with cached input at $0.30/M.
Verdict
For prompts up to 200K, Gemini charges $2.00/M input and $12.00/M output against Sonnet's $3.00 and $15.00, so it wins both base rows by 33% on input and 20% on output. Gemini also wins cached input at $0.20/M against Sonnet's $0.30/M. Both discounts are 90%. Sonnet's edges are its tool use and instruction following, while Gemini offers a 1M context window against Sonnet's 500K. Above 200K, Gemini's standard rates rise to $4.00/M input and $18.00/M output, so check the long-context tier before choosing.
Which should you pick?
Choose Gemini 3.1 Pro (<=200k)
Long-context processing up to 1M tokens, native multimodal input, and lower costs for prompts up to 200K.
Choose Claude Sonnet 4.6
State-of-the-art multi-file code editing, tool use, and agentic loops.