Compare · ModelsLive · 2 picked · head to head
Claude Haiku 5.5 vs Kimi K2.7 Code
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Claude Haiku 5.5 wins on 13/14 benchmarks
Claude Haiku 5.5 wins 13 of 14 shared benchmarks. Leads in speed · agentic · knowledge.
Category leads
speed·Claude Haiku 5.5agentic·Claude Haiku 5.5knowledge·Claude Haiku 5.5coding·Claude Haiku 5.5math·Claude Haiku 5.5
Hype vs Reality
Attention vs performance
Claude Haiku 5.5
#109 by perf·#15 by attention
Kimi K2.7 Code
#77 by perf·#18 by attention
Best value
Claude Haiku 5.5
6.3x better value than Kimi K2.7 Code
Claude Haiku 5.5
175.3 pts/$
$0.30/M
Kimi K2.7 Code
27.8 pts/$
$2.01/M
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
Moonshot AI
$50.0B·Tier 1
Head to head
14 benchmarks · 2 models
Claude Haiku 5.5Kimi K2.7 Code
Artificial Analysis · CritPt
Claude Haiku 5.5 leads by +8.9
Claude Haiku 5.5
18.9
Kimi K2.7 Code
10.0
Artificial Analysis · GDPval
Claude Haiku 5.5 leads by +29.0
Claude Haiku 5.5
55.9
Kimi K2.7 Code
26.9
Artificial Analysis · Humanity's Last Exam
Claude Haiku 5.5 leads by +9.4
Claude Haiku 5.5
44.4
Kimi K2.7 Code
35.0
Artificial Analysis · Long Context Reasoning
Claude Haiku 5.5 leads by +3.4
Claude Haiku 5.5
82.7
Kimi K2.7 Code
79.3
Artificial Analysis · Quality Index
Claude Haiku 5.5 leads by +17.6
Claude Haiku 5.5
43.4
Kimi K2.7 Code
25.8
Artificial Analysis · SciCode
Claude Haiku 5.5 leads by +7.2
Claude Haiku 5.5
55.0
Kimi K2.7 Code
47.8
APEX-Agents
Claude Haiku 5.5 leads by +18.5
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Claude Haiku 5.5
56.1
Kimi K2.7 Code
37.6
Chess Puzzles
Claude Haiku 5.5 leads by +43.1
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Claude Haiku 5.5
52.6
Kimi K2.7 Code
9.5
Frontiercode
Claude Haiku 5.5 leads by +16.3
Claude Haiku 5.5
46.4
Kimi K2.7 Code
30.1
FrontierMath-Tier-4-v2-Private
Claude Haiku 5.5 leads by +34.1
Claude Haiku 5.5
46.3
Kimi K2.7 Code
12.2
FrontierMath-Tiers-1-3-v2-Private
Claude Haiku 5.5 leads by +21.1
Claude Haiku 5.5
75.1
Kimi K2.7 Code
54.0
GPQA diamond
Claude Haiku 5.5 leads by +2.4
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude Haiku 5.5
86.2
Kimi K2.7 Code
83.8
OTIS Mock AIME 2024-2025
Claude Haiku 5.5 leads by +3.3
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Haiku 5.5
98.9
Kimi K2.7 Code
95.5
SimpleQA Verified
Kimi K2.7 Code leads by +12.7
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Claude Haiku 5.5
23.8
Kimi K2.7 Code
36.5
Full benchmark table
| Benchmark | Claude Haiku 5.5 | Kimi K2.7 Code |
|---|---|---|
Artificial Analysis · CritPt | 18.9 | 10.0 |
Artificial Analysis · GDPval | 55.9 | 26.9 |
Artificial Analysis · Humanity's Last Exam | 44.4 | 35.0 |
Artificial Analysis · Long Context Reasoning | 82.7 | 79.3 |
Artificial Analysis · Quality Index | 43.4 | 25.8 |
Artificial Analysis · SciCode | 55.0 | 47.8 |
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 56.1 | 37.6 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 52.6 | 9.5 |
Frontiercode | 46.4 | 30.1 |
FrontierMath-Tier-4-v2-Private | 46.3 | 12.2 |
FrontierMath-Tiers-1-3-v2-Private | 75.1 | 54.0 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 86.2 | 83.8 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 98.9 | 95.5 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 23.8 | 36.5 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.10 | $0.50 | 1.0M tokens (~500 books) | $2.00 | |
| $0.67 | $3.35 | 262K tokens (~131 books) | $13.41 |