Compare · ModelsLive · 2 picked · head to head
Kimi K2.5 vs Kimi K2.7 Code
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Kimi K2.7 Code wins on 8/10 benchmarks
Kimi K2.7 Code wins 8 of 10 shared benchmarks. Leads in agentic · knowledge · math.
Category leads
speed·Kimi K2.5agentic·Kimi K2.7 Codeknowledge·Kimi K2.7 Codemath·Kimi K2.7 Codereasoning·Kimi K2.7 Codecoding·Kimi K2.7 Code
Hype vs Reality
Attention vs performance
Kimi K2.5
#122 by perf·#17 by attention
Kimi K2.7 Code
#69 by perf·#17 by attention
Best value
Kimi K2.5
1.2x better value than Kimi K2.7 Code
Kimi K2.5
37.0 pts/$
$1.35/M
Kimi K2.7 Code
30.5 pts/$
$1.84/M
Vendor risk
Who is behind the model
Moonshot AI
$18.0B·Tier 1
Moonshot AI
$18.0B·Tier 1
Head to head
10 benchmarks · 2 models
Kimi K2.5Kimi K2.7 Code
Artificial Analysis · Agentic Index
Kimi K2.5 leads by +29.3
Artificial Analysis Agentic Index · a composite score measuring how well a model performs in agentic workflows · multi-step tool use, planning, error recovery, and autonomous task completion. Aggregates results from multiple agentic benchmarks including SWE-bench, tool-use tests, and planning evaluations. The canonical single-number metric for "how good is this model as an agent?"
Kimi K2.5
58.9
Kimi K2.7 Code
29.6
Artificial Analysis · Coding Index
Kimi K2.7 Code leads by +21.2
Artificial Analysis Coding Index · a composite score that aggregates performance across multiple coding benchmarks into a single index. Tracks code generation quality, debugging ability, multi-language competence, and real-world software engineering tasks. Used by Artificial Analysis to rank model coding capability in a normalized, comparable format. Useful for developers choosing between models for coding-heavy workloads.
Kimi K2.5
39.5
Kimi K2.7 Code
60.8
Artificial Analysis · Quality Index
Kimi K2.5 leads by +21.0
Kimi K2.5
46.8
Kimi K2.7 Code
25.8
APEX-Agents
Kimi K2.7 Code leads by +23.2
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Kimi K2.5
14.4
Kimi K2.7 Code
37.6
Chess Puzzles
Kimi K2.7 Code leads by +9.5
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Kimi K2.5
7.4
Kimi K2.7 Code
16.9
GPQA diamond
Kimi K2.7 Code leads by +0.4
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Kimi K2.5
83.5
Kimi K2.7 Code
83.8
OTIS Mock AIME 2024-2025
Kimi K2.7 Code leads by +3.4
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Kimi K2.5
92.2
Kimi K2.7 Code
95.5
SimpleBench
Kimi K2.7 Code leads by +13.3
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
Kimi K2.5
36.2
Kimi K2.7 Code
49.5
SimpleQA Verified
Kimi K2.7 Code leads by +2.2
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Kimi K2.5
34.3
Kimi K2.7 Code
36.5
WeirdML
Kimi K2.7 Code leads by +8.5
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Kimi K2.5
45.6
Kimi K2.7 Code
54.1
Full benchmark table
| Benchmark | Kimi K2.5 | Kimi K2.7 Code |
|---|---|---|
Artificial Analysis · Agentic Index Artificial Analysis Agentic Index · a composite score measuring how well a model performs in agentic workflows · multi-step tool use, planning, error recovery, and autonomous task completion. Aggregates results from multiple agentic benchmarks including SWE-bench, tool-use tests, and planning evaluations. The canonical single-number metric for "how good is this model as an agent?" | 58.9 | 29.6 |
Artificial Analysis · Coding Index Artificial Analysis Coding Index · a composite score that aggregates performance across multiple coding benchmarks into a single index. Tracks code generation quality, debugging ability, multi-language competence, and real-world software engineering tasks. Used by Artificial Analysis to rank model coding capability in a normalized, comparable format. Useful for developers choosing between models for coding-heavy workloads. | 39.5 | 60.8 |
Artificial Analysis · Quality Index | 46.8 | 25.8 |
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 14.4 | 37.6 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 7.4 | 16.9 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 83.5 | 83.8 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 92.2 | 95.5 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 36.2 | 49.5 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 34.3 | 36.5 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 45.6 | 54.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.45 | $2.25 | 262K tokens (~131 books) | $9.00 | |
| $0.61 | $3.07 | 262K tokens (~131 books) | $12.26 |