Claude pelican comparison: Fable, Opus and Sonnet

Prompt: “Generate an SVG of a pelican riding a bicycle”. Five thinking-effort levels, shown from max to low.

Prices per 1M input/output tokens: Claude Fable 5.1 $10/$50; Claude Opus 5.5 $4/$20; Claude Opus 5 $5/$25; Claude Sonnet 5 $2/$10. Cached input: Fable 5.1 $0.25/M; Opus 5.5 $0.20/M.

Calculated costs use the supplied prices and logged token counts, including thinking tokens in the output total. Input is charged at the uncached rate because the source logs do not specify cached usage. Costs are rounded to two decimal places in cents. Fable runs: September 1, 2026; all other runs: September 22, 2026.

EffortClaude Fable 5.1Claude Opus 5.5Claude Opus 5Claude Sonnet 5
max
Claude Fable 5.1, thinking effort max
27 input tokens, 65,927 output = 329.66 cents
59,725 thinking tokens.
No SVG returnedThe response is empty. All 128,000 output tokens were thinking tokens.
27 input tokens, 128,000 output = 256.01 cents
128,000 thinking tokens.
Claude Opus 5, thinking effort max
25 input tokens, 13,207 output = 33.03 cents
8,976 thinking tokens.
Claude Sonnet 5, thinking effort max
25 input tokens, 69,340 output = 69.35 cents
66,485 thinking tokens.
xhigh
Claude Fable 5.1, thinking effort xhigh
27 input tokens, 36,767 output = 183.86 cents
33,128 thinking tokens.
Claude Opus 5.5, thinking effort xhigh
27 input tokens, 11,977 output = 23.96 cents
8,821 thinking tokens.
Claude Opus 5, thinking effort xhigh
25 input tokens, 7,295 output = 18.25 cents
3,372 thinking tokens.
Claude Sonnet 5, thinking effort xhigh
25 input tokens, 2,503 output = 2.51 cents
0 thinking tokens.
high
Claude Fable 5.1, thinking effort high
27 input tokens, 2,612 output = 13.09 cents
305 thinking tokens.
Claude Opus 5.5, thinking effort high
27 input tokens, 3,366 output = 6.74 cents
329 thinking tokens.
Claude Opus 5, thinking effort high
25 input tokens, 6,029 output = 15.09 cents
1,792 thinking tokens.
Claude Sonnet 5, thinking effort high
25 input tokens, 1,958 output = 1.96 cents
32 thinking tokens.
medium
Claude Fable 5.1, thinking effort medium
27 input tokens, 1,977 output = 9.91 cents
0 thinking tokens.
Claude Opus 5.5, thinking effort medium
27 input tokens, 2,941 output = 5.89 cents
220 thinking tokens.
Claude Opus 5, thinking effort medium
25 input tokens, 3,496 output = 8.75 cents
586 thinking tokens.
Claude Sonnet 5, thinking effort medium
25 input tokens, 2,168 output = 2.17 cents
0 thinking tokens.
low
Claude Fable 5.1, thinking effort low
27 input tokens, 1,998 output = 10.02 cents
0 thinking tokens.
Claude Opus 5.5, thinking effort low
27 input tokens, 1,891 output = 3.79 cents
0 thinking tokens.
Claude Opus 5, thinking effort low
25 input tokens, 1,780 output = 4.46 cents
22 thinking tokens.
Claude Sonnet 5, thinking effort low
25 input tokens, 1,370 output = 1.38 cents
0 thinking tokens.

Fable 5.1 thinking-token counts are inferred: logged output tokens minus the full response text counted by the Anthropic token-counting API, with a six-token message-overhead correction calibrated against the low and medium runs, which contain no reasoning blocks. The API provides estimates; these are not recorded usage breakdowns. Calculation details.