code
Apr 02, 2026CODE-025Implement an LRU cache with per-key TTL (time-to-live) support. Requirements: O(1) get/put, thread-safe, lazy expiration (don't use background threads), configurable max size, eviction callback, and cache hit/miss statistics. Include comprehensive tests.
Winner
Gemini 3 Flash Preview
Google
8.34
WINNER SCORE
matrix avg: 6.43
10×10 Judgment Matrix · 78 judgments
OPEN DATA
| Judge ↓ / Respondent → | DeepSeek V4 | GPT-OSS-120B | MiMo-V2-Flash | GPT-5.4 | Claude Opus 4.6 | Gemini 3.1 Pro | Claude Sonnet 4.6 | Grok 4.20 | Gemini 3 | MiniMax M2.5 |
|---|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 | — | 9.2 | 9.6 | 9.6 | 8.8 | · | 8.8 | 9.2 | 9.6 | · |
| GPT-OSS-120B | 7.5 | — | 6.5 | 7.7 | 4.8 | 2.9 | 3.2 | 7.0 | 8.8 | · |
| MiMo-V2-Flash | 7.8 | 8.6 | — | 8.6 | 8.6 | · | 7.6 | 8.2 | 9.0 | · |
| GPT-5.4 | 4.3 | 2.0 | 4.3 | — | 3.5 | 1.6 | 3.0 | 4.4 | 6.8 | · |
| Claude Opus 4.6 | 6.0 | 4.9 | 6.5 | 6.5 | — | 1.2 | 5.5 | 5.6 | 7.8 | · |
| Gemini 3.1 Pro | 6.2 | 5.2 | 6.2 | 5.5 | 5.8 | — | 6.5 | 4.5 | 9.0 | · |
| Claude Sonnet 4.6 | 6.8 | 6.0 | 7.0 | 8.2 | 6.7 | 1.9 | — | 6.8 | 7.5 | · |
| Grok 4.20 | 6.0 | 5.8 | 6.2 | 8.6 | 6.0 | 3.6 | 6.4 | — | 8.0 | · |
| Gemini 3 | 8.0 | 7.8 | 9.0 | 9.6 | 9.0 | 4.3 | 8.4 | 8.6 | — | · |
| MiniMax M2.5 | · | 4.7 | 6.9 | 6.8 | 4.5 | 3.6 | 5.7 | 7.5 | 8.6 | — |