Fugu Ultra vs GPT-5.5
Prices checked
GPT-5.5 takes 2 of 5 rounds against Fugu Ultra. It is also the cheaper of the two to run, at $10.23 per million tokens against $11.25.
Sakana AIFugu Ultra
1M context · tool calling · reads images
$11.25/M tokensSakana AI's own price · $4.80 per 1,000 pages
OpenAIWins 2 roundsGPT-5.5
1.1M context · tool calling · reads images
$10.23/M tokensCheapest price, at Poe · $4.36 per 1,000 pages
Maker's price in (lower wins)
$5/M
$5/M
Maker's price out (lower wins)
$30/M
$30/M
Cheapest price (lower wins)
none below the maker
$10.23/M at Poe
Maker's price, blended (lower wins)
$11.25/M
$11.25/M
Cost per 1,000 pages (lower wins)
$4.80
$4.36
Context
1M tokens
1.1M tokens
Open weights
not published
no
Tool calling
yes
yes
Structured output
yes
yes
Reads images
yes
yes
Reasoning
yes
yes
Released
24 Jun 2026
23 Apr 2026
Providers selling it
13
53
Price we measure against
the maker's own
the maker's own
Trains on your prompts (no is better)
no provider to check
not published
Providers selling both · 11
- CommandCode$11.25 / $11.25
Kilo Gateway$11.25 / $11.25
LLM Gateway$11.25 / $11.25
LLMTR$11.25 / $11.25
Merge Gateway$11.25 / $11.25
NanoGPT$11.25 / $11.25
Nous Portal$11.25 / $11.25
OpenRouter$11.25 / $11.25
Fugu Ultra, then GPT-5.5, per million tokens; the 8 cheapest of 11.
Go further
Add a third or fourth model, or change the weighting, in the tool.
Frequently asked
- Which is cheaper, Fugu Ultra or GPT-5.5?
- GPT-5.5, at $10.23 per million tokens against $11.25 for Fugu Ultra, blended 3 in : 1 out, read on 31 Oct 2026.
- Where is each one cheapest?
- Fugu Ultra: nobody sells it below the maker's own price with a stated reason, so $11.25 is the number to beat. GPT-5.5: $10.23 at Poe.
- Can I buy both from one provider?
- Yes. 11 providers sell both: CommandCode, Kilo Gateway, LLM Gateway, LLMTR, Merge Gateway, NanoGPT, and more. The cheapest of them for the two together is CommandCode.
- Which holds more context?
- GPT-5.5, at 1.1M tokens against 1M.
Prices exclude tax, read 31 Oct 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