Data points: 80
Model compare
The readout for Kimi K2.5 and Kimi K2 Thinking, before the detailed comparison sheet.
Weighted outcome: Kimi K2 Thinking. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.
Decision read
Kimi K2 Thinking
Kimi K2 Thinking has the higher weighted result; Model A / B score 46.7 to 53.3.
Evidence depth
80 data points
Includes 7 benchmark rows, 0 audit samples, and 7 provider examples.
Selection signal
Start with Kimi K2 Thinking
The charts below split 14 high-signal samples across speed, scores, and audit health.
Switch either side of this report to compare another model with the same LMSpeed data pipeline.
Select a different model to open a new comparison URL.
This report only uses LMSpeed data for Kimi K2.5 and Kimi K2 Thinking: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.
| Model compare | Kimi K2.5 | Kimi K2 Thinking |
|---|---|---|
| Overall leader | Contender | Leading |
| Weighted overall score | 46.7 pts | 53.3 pts |
| Benchmark category leads | 1 categories | 2 categories |
| Operational advantages | Cheapest input price, Average speed, First-token latency, Free providers, Provider coverage | No data |
| Context window | 262.1K tokens | 262.1K tokens |
| Max output | 262.1K tokens | 100.4K tokens |
| Modalities | Input TextImage Output Text | Input |
The overall result weights benchmark capability categories at 80% and price, API speed/latency, and availability at 20%. Recent test volume does not affect the winner, and missing benchmark categories are excluded.
| Model compare | Kimi K2.5 | Kimi K2 Thinking |
|---|---|---|
| Developer | MoonshotAI | MoonshotAI |
| Released | Jan 2026 | Nov 2025 |
| Parameters | No data | No data |
| Tokenizer | Other | Other |
| Knowledge cutoff | No data | No data |
| OpenRouter ID | moonshotai/kimi-k2.5 | moonshotai/kimi-k2-thinking |
| References | No data | No data |
This report only uses LMSpeed data for Kimi K2.5 and Kimi K2 Thinking: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.
Kimi K2.5
Kimi K2.5 is stronger in benchmark categories (Reasoning) and operational dimensions (Cheapest input price, Average speed, First-token latency, Free providers, Provider coverage).
Kimi K2 Thinking
Kimi K2 Thinking leads these benchmark categories: Coding, Math.
Third-party benchmark profile synced into LMSpeed; only metrics available for both models are shown.
Compare benchmark category scores on a 0-100 scale. Select a category to inspect the gap.
Avg. score
Kimi K2.5
49.7
Avg. score
Kimi K2 Thinking
55.7
Kimi K2.5
Kimi K2 Thinking leads by 9.0
Kimi K2.5 leads by 0.2
Kimi K2.5
Kimi K2 Thinking leads by 4.4
Kimi K2.5
Kimi K2.5
Kimi K2.5
Metric-level scores with benchmark source, rank depth, confidence, error, and evaluation date where available.
Kimi K2.5
$3.00/M
Rank #85/162 · confidence 4
Kimi K2 Thinking
winner$2.50/M
Rank #75/162 · confidence 4
Kimi K2.5
$1.20/M
Rank #85/162 · confidence 4
Kimi K2 Thinking
winner$1.07/M
Rank #78/162 · confidence 4
Kimi K2.5
$0.600/M
Rank #80/162 · confidence 4
Kimi K2 Thinking
$0.600/M
Rank #80/162 · confidence 4
Kimi K2.5
winner87.1%
Rank #11/125 · confidence 4 · eval date 2026-02-01
Kimi K2 Thinking
84.8%
Rank #28/125 · confidence 4
Kimi K2.5
12.3%
Rank #84/187 · confidence 4
Kimi K2 Thinking
winner22.3%
Rank #49/187 · confidence 4
Kimi K2.5
78.9%
Rank #79/188 · confidence 4
Kimi K2 Thinking
winner83.8%
Rank #55/188 · confidence 4
Kimi K2.5
39.6%
Rank #76/185 · confidence 4
Kimi K2 Thinking
winner42.4%
Rank #49/185 · confidence 4
Latest completed audits from shared providers, with four safety and integrity score groups plus report links.
| Provider | Kimi K2.5 | Kimi K2 Thinking |
|---|---|---|
| No completed audits are available from shared providers yet. | ||
Speed aggregates and input/output pricing share each provider row for real API selection and migration cost checks.
| Provider | Kimi K2.5 | Kimi K2 Thinking |
|---|---|---|
150 tests | Kimi K2.5 speed / latency 45 tok/s / 2058ms input / output No data | Kimi K2 Thinking speed / latency 34 tok/s / 16242ms input / output No data |
85 tests | Kimi K2.5 speed / latency 48 tok/s / 18528ms input / output No data | Kimi K2 Thinking speed / latency 41 tok/s / 19451ms input / output No data |
30 tests | Kimi K2.5 speed / latency 18 tok/s / 25450ms input / output No data | Kimi K2 Thinking speed / latency N/A / N/A input / output No data |
25 tests | Kimi K2.5 speed / latency 38 tok/s / 22910ms input / output No data | Kimi K2 Thinking speed / latency N/A / N/A input / output No data |
10 tests | Kimi K2.5 speed / latency 45 tok/s / 14302ms input / output No data | Kimi K2 Thinking speed / latency N/A / N/A input / output No data |
Kimi K2.5 kimi-k2.5 speed / latency No data input / output $0.041/M/$0.205/M | Kimi K2 Thinking kimi-k2-thinking speed / latency No data input / output $0.027/M | |
Kimi K2.5 kimi-k2.5 speed / latency No data input / output $0.010/request | Kimi K2 Thinking kimi-k2-thinking-tidy speed / latency No data input / output $0.010/request |
Weighted outcome: Kimi K2 Thinking. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.
Output
| Features | Text inputImage inputText outputTool callingStructured outputsJSON modeReasoning | Text inputText outputTool callingStructured outputsJSON modeReasoning |
|---|
