Side-by-side comparison · models
Gemma 4 31B vs Kimi K2.5 TEE
What we have checked about each, line by line. Every value shows where it came from and when we last confirmed it.
Models
Gemma 4 31B
Open Gemma instruction model for efficient chat and self-hosted deployments.
Models
Kimi K2.5 TEE
Kimi model for long-context chat, coding, and agentic reasoning.
Checked facts
Line by line
| Measure | Gemma 4 31B | Kimi K2.5 TEE |
|---|---|---|
| Input price | Not checked yet | $0.532 per million tokens |
| Output price | Not checked yet | $3.33 per million tokens |
| Price on amazon-bedrock | Not checked yet | $0.600 per million input tokens |
| Price on amazon-bedrock | Not checked yet | $3.00 per million output tokens |
| Price on atlas-cloud | Not checked yet | $0.200 per million cached input tokens |
| Price on atlas-cloud | Not checked yet | $0.490 per million input tokens |
| Price on atlas-cloud | Not checked yet | $2.50 per million output tokens |
| Price on novita | Not checked yet | $0.095 per million cached input tokens |
| Price on novita | Not checked yet | $0.570 per million input tokens |
| Price on novita | Not checked yet | $2.85 per million output tokens |
| Price on siliconflow | Not checked yet | $0.150 per million cached input tokens |
| Price on siliconflow | Not checked yet | $0.500 per million input tokens |
| Price on siliconflow | Not checked yet | $2.50 per million output tokens |
| Price on venice | Not checked yet | $0.209 per million cached input tokens |
| Price on venice | Not checked yet | $0.532 per million input tokens |
| Price on venice | Not checked yet | $3.33 per million output tokens |
| Latest version | Not checked yet | moonshotai/kimi-k2.5-0127 |
| Time to first token (ms) | 1,100 | 1,780 |
| Output speed (tokens a second) | 34.7 | 45.4 |
| Time to first token (ms) on amazon-bedrock | Not checked yet | 1,140 |
| Time to first token (ms) on atlas-cloud | Not checked yet | 817 |
| Time to first token (ms) on novita | Not checked yet | 1,374 |
| Time to first token (ms) on siliconflow | Not checked yet | 1,549 |
| Time to first token (ms) on venice | Not checked yet | 1,011 |
| Output speed (tokens a second) on amazon-bedrock | Not checked yet | 52 |
| Output speed (tokens a second) on atlas-cloud | Not checked yet | 45 |
| Output speed (tokens a second) on novita | Not checked yet | 38 |
| Output speed (tokens a second) on siliconflow | Not checked yet | 22 |
| Output speed (tokens a second) on venice | Not checked yet | 22 |
| Tokens served on OpenRouter in a day | Not checked yet | 3.8B (for 10 Nov 2026) |
| Tool calls on OpenRouter in a day | Not checked yet | 46.7K (for 10 Nov 2026) |
| Artificial Analysis Intelligence Index | Not checked yet | 23.5 |
| Artificial Analysis Coding Index | Not checked yet | 46.8 |
| LMArena text rating (Elo) | Not checked yet | 1,445 |
| LMArena web development rating (Elo) | Not checked yet | 1,436 |
| LMArena vision rating (Elo) | Not checked yet | 1,269 |
| Design Arena rating (Elo) | Not checked yet | 1,245 |
A line appears only when at least one record has a value we confirmed from a source. Values we have not confirmed are left out, never guessed. Records of different types share fewer lines.
Add a third
Compare with one more record
How this works
Share it, or build your own
This page works for any two or three records in the directory, of any type. The link stays the same whichever order you name them in, so you can share it as it is. Change the records in the address, for example /directories/vs/claude-code-vs-cursor, or start from our editorial comparisons.