Claude Fable 5
VS
Kimi K2.7 Code
Claude Fable 5 vs Kimi K2.7 Code: which wins at real work?
22 task areas · same graded test runs · rank comparison only, so 0–100 and Elo collections never mix raw scores.
Claude Fable 5 wins 15 of 22 task areas we tested; Kimi K2.7 Code takes 7. Kimi K2.7 Code costs 14.2× less per token ($4.24 vs $60 per 1M).
Kimi K2.7 Code costs 14.2× less per token ($4.24 vs $60 per 1M).
Task by task
| Task area | Claude Fable 5 | Kimi K2.7 Code | Winner |
|---|---|---|---|
| Product & Project Management |
#76
/ 110
Strong
|
#10
/ 110
Excellent
|
Kimi K2.7 Code |
| Chef / Home Cooking |
#3
/ 126
Strong
|
#58
/ 126
Usable
|
Claude Fable 5 |
| Content & Brand |
#1
/ 124
Excellent
|
#48
/ 124
Strong
|
Claude Fable 5 |
| AI Strategy |
#1
/ 126
Strong
|
#44
/ 126
Strong
|
Claude Fable 5 |
| Creative & Comedy | #1 / 110 | #44 / 110 | Claude Fable 5 |
| Sales |
#5
/ 110
Strong
|
#48
/ 110
Usable
|
Claude Fable 5 |
| Coding |
#11
/ 115
Excellent
|
#41
/ 115
Strong
|
Claude Fable 5 |
| Translation & Localization |
#18
/ 110
Excellent
|
#48
/ 110
Excellent
|
Claude Fable 5 |
| Data & Analytics |
#3
/ 110
Excellent
|
#31
/ 110
Excellent
|
Claude Fable 5 |
| Summarization & Meeting Notes |
#24
/ 110
Excellent
|
#47
/ 110
Excellent
|
Claude Fable 5 |
| Training & Education |
#16
/ 110
Excellent
|
#39
/ 110
Excellent
|
Claude Fable 5 |
| Structured Output |
#68
/ 113
Strong
|
#46
/ 113
Excellent
|
Kimi K2.7 Code |
| Research & Competitive Analysis |
#3
/ 110
Excellent
|
#20
/ 110
Excellent
|
Claude Fable 5 |
| Executive Assistant |
#22
/ 112
Strong
|
#6
/ 112
Strong
|
Kimi K2.7 Code |
| Landing Pages |
#1
/ 72
Strong
|
#17
/ 72
Strong
|
Claude Fable 5 |
| Presentations & Decks |
#20
/ 110
Excellent
|
#5
/ 110
Excellent
|
Kimi K2.7 Code |
| Legal & HR |
#1
/ 110
Excellent
|
#14
/ 110
Excellent
|
Claude Fable 5 |
| Investor & Pitch |
#3
/ 66
Strong
|
#15
/ 66
Strong
|
Claude Fable 5 |
| Knowledge & Docs |
#27
/ 110
Strong
|
#16
/ 110
Excellent
|
Kimi K2.7 Code |
| Customer Support |
#1
/ 113
Excellent
|
#10
/ 113
Strong
|
Claude Fable 5 |
| RAG, Safety & Grounding |
#30
/ 113
Excellent
|
#21
/ 113
Excellent
|
Kimi K2.7 Code |
| Frontend & Landing Pages |
#58
/ 109
Needs editing
|
#54
/ 109
Needs editing
|
Kimi K2.7 Code |
Rank = position among every model config we tested in that task area (lower is better). Sorted by biggest gap first.
Same task, both models — judged
Both models answered the same test case; an independent judge graded each. Quotes are the judge's actual rationale.
Content & Brand
AI consulting LinkedIn post (Generic Copy Index)
“The response is exceptional. It perfectly adheres to all constraints, including the strict word count and negative constraints. The voice is spot-on for the target audience, the example is highly specific and realistic, and the advice is practical and actionable. It reads like a genuine, high-quality post from an industry expert.”
“The response failed the hard length constraint of 150-220 words as it was only 142 words long.”
Frequently asked
Is Claude Fable 5 better than Kimi K2.7 Code?
Across 22 task areas we benchmarked, Claude Fable 5 ranks higher in 15 and Kimi K2.7 Code in 7.
Which is cheaper, Claude Fable 5 or Kimi K2.7 Code?
Kimi K2.7 Code costs 14.2× less per token ($4.24 vs $60 per 1M).
What is Claude Fable 5 better at?
Claude Fable 5 out-ranks Kimi K2.7 Code at Chef / Home Cooking, Content & Brand, AI Strategy.
What is Kimi K2.7 Code better at?
Kimi K2.7 Code out-ranks Claude Fable 5 at Product & Project Management, Structured Output, Executive Assistant.
More comparisons
This page is Spring Prompt, running in public
We just did this for every model. Do it for your prompt.
The rankings above come from running real tasks through real models and scoring every output. Spring Prompt is that same engine — pointed at your prompt, your test cases, and your definition of good.
- Generate test cases from your prompt — no eval set required to start.
- Compare models side by side with quality, cost and latency in one matrix.
- Optimise the winner until the scores say it's ready to ship.
Prompt × model results
12 test cases · 3 evals