Claude Sonnet 5.5 vs Gemini 4 Carbon
Prices checked
Gemini 4 Carbon takes 1 of 5 rounds against Claude Sonnet 5.5. Both cost $4 per million tokens to run.
Claude Sonnet 5.5
1M context · tool calling · reads images
$4/M tokensAnthropic's own price · $1.80 per 1,000 pages
Gemini 4 Carbon
1M context · tool calling · reads images
$4/M tokensGoogle's own price · $1.80 per 1,000 pages
Maker's price in (lower wins)
$2/M
$2/M
Maker's price out (lower wins)
$10/M
$10/M
Cheapest price (lower wins)
none below the maker
none below the maker
Maker's price, blended (lower wins)
$4/M
$4/M
Cost per 1,000 pages (lower wins)
$1.80
$1.80
Context
1M tokens
1M tokens
Open weights
no
no
Tool calling
yes
yes
Structured output
yes
yes
Reads images
yes
yes
Reasoning
yes
yes
Released
28 Sept 2026
10 Nov 2026
Providers selling it
44
26
Price we measure against
the maker's own
the maker's own
Trains on your prompts (no is better)
no provider to check
no provider to check
Providers selling both · 24
NanoGPT$4 / $1.53
Cloudflare AI Gateway$4 / $4
CrossModel$4 / $4
Eden AI$4 / $4
FastRouter$4 / $4
GitHub Copilot$4 / $4
Google$4 / $4
Google Vertex AI$4 / $4
Claude Sonnet 5.5, then Gemini 4 Carbon, per million tokens; the 8 cheapest of 24.
Go further
Add a third or fourth model, or change the weighting, in the tool.
Frequently asked
- Which is cheaper, Claude Sonnet 5.5 or Gemini 4 Carbon?
- Neither: both run at $4 per million tokens, blended 3 in : 1 out, read on 13 Nov 2026.
- Where is each one cheapest?
- Claude Sonnet 5.5: nobody sells it below the maker's own price with a stated reason, so $4 is the number to beat. Gemini 4 Carbon: nobody sells it below the maker's own price with a stated reason, so $4 is the number to beat.
- Can I buy both from one provider?
- Yes. 24 providers sell both: NanoGPT, Cloudflare AI Gateway, CrossModel, Eden AI, FastRouter, GitHub Copilot, and more. The cheapest of them for the two together is NanoGPT.
- Which holds more context?
- Gemini 4 Carbon, at 1M tokens against 1M.
Prices exclude tax, read 13 Nov 2026 from each provider's own catalog. They are per M tokens, blended 3 in : 1 out, and a page is 600 tokens in and 60 out. Bars are scaled to the larger value in each row.
Related comparisons