head-to-head

Anthropic Claude Haiku Latest vs MoonshotAI: Kimi K2.6

Side-by-side comparison of specs, pricing, benchmark scores, and task rankings. Updated 2026-08-06.

	Anthropic Claude Haiku Latest	MoonshotAI: Kimi K2.6
Vendor	~anthropic	moonshotai
Quality Score	100	100
Benchmark Score	-	79.3
Input Price	$1.00/M	$0.59/M
Output Price	$5.00/M	$2.48/M
Context Window	200,000	262,144
Max Output	64,000	262,144
Tool Calling	✓	✓
Structured Output	✓	✓
Reasoning Mode	✓	✓
Vision	✓	✓
Audio	-	-
Benchmark Scores
ai_index	-	73.0
ai_index_agentic	-	49.9
ai_index_coding	-	100.0
eqbench	-	76.6

Who wins by task?

Task	Anthropic Claude Haiku Latest	MoonshotAI: Kimi K2.6
SQL Generation	129	169
Code Review	124	165
Code Completion	114	132
Code Refactoring	124	161
Bug Fixing	128	176
Unit Test Generation	120	153
Code Documentation	122	142
Regex Writing	118	136
CI/CD Pipelines	116	143
Frontend Component Design	122	146
Data Analysis	124	167
CSV / Spreadsheet Cleanup	125	152
ETL Scripting	120	152
JSON Extraction	122	147
Bulk Data Labeling	120	134
OCR / Document Parsing	127	145
Table Extraction from PDFs	127	145
Long-Document Summarization	125	158
Short-Form Summarization	114	132
Blog Post Writing	117	140

Scores reflect capability match + benchmark data + pricing for each task. Methodology →