MoonshotAI: Kimi K3 vs MoonshotAI: Kimi K2.7 Code
Side-by-side comparison of specs, pricing, benchmark scores, and task rankings. Updated 2026-11-12.
| MoonshotAI: Kimi K3 | MoonshotAI: Kimi K2.7 Code | |
|---|---|---|
| Vendor | moonshotai | moonshotai |
| Quality Score | 100 | 100 |
| Input Price | $0.19/M | $0.60/M |
| Output Price | $16.00/M | $3.50/M |
| Context Window | 1,048,576 | 262,144 |
| Max Output | 943,718 | 235,929 |
| Tool Calling | ✓ | ✓ |
| Structured Output | ✓ | ✓ |
| Reasoning Mode | ✓ | ✓ |
| Vision | ✓ | ✓ |
| Audio | - | - |
The verdict
For a sample job of 1 million input tokens and 250,000 output tokens, Kimi K2.7 Code costs $1.48 and Kimi K3 costs $4.19, so Kimi K2.7 Code is about 2.8 times cheaper for the same work.
Kimi K3 reads up to 1,048,576 tokens at once, roughly 1,573 pages of text, against about 393 pages for Kimi K2.7 Code.
Only Kimi K3 offers video input.
Across the 18 tasks where they differ, Kimi K3 scores higher on 14 (SQL Generation, Code Review, Code Completion and others), Kimi K2.7 Code on 4 (Regex Writing, JSON Extraction, Bulk Data Labeling and others).
Kimi K3 was released in July 2026 and Kimi K2.7 Code in June 2026.
Pick Kimi K3 if you need long documents, video input or the stronger all-round task scores; pick Kimi K2.7 Code if you need lower cost.
Costs use current list prices per million tokens; page counts assume about 0.75 words per token and 500 words per page.
Who wins by task?
| Task | MoonshotAI: Kimi K3 | MoonshotAI: Kimi K2.7 Code |
|---|---|---|
| SQL Generation | 132 | 130 |
| Code Review | 132 | 126 |
| Code Completion | 128 | 128 |
| Code Refactoring | 136 | 127 |
| Bug Fixing | 136 | 130 |
| Unit Test Generation | 124 | 121 |
| Code Documentation | 128 | 125 |
| Regex Writing | 116 | 119 |
| CI/CD Pipelines | 120 | 117 |
| CSV / Spreadsheet Cleanup | 132 | 128 |
| ETL Scripting | 128 | 122 |
| JSON Extraction | 128 | 131 |
| Bulk Data Labeling | 125 | 128 |
| OCR / Document Parsing | 131 | 128 |
| Table Extraction from PDFs | 131 | 128 |
| Long-Document Summarization | 136 | 129 |
| Short-Form Summarization | 120 | 123 |
| Blog Post Writing | 120 | 118 |
Scores reflect capability match + benchmark data + pricing for each task. Methodology →