LOADING COMPARISON
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
HEAD TO HEAD
Qwen2.5 Coder 32B Instruct vs Codestral vs Qwen3.8 27B
Community vibe score against API price per 1M tokens. Up and left wins: loved and cheap.
WHAT PEOPLE SAY
Codestral
LOVE
Popular for coding help
Qwen3.8 27B
LOVE
Popular for coding help
COMPLAINT
Refuses or lectures more than people like
VALUE FOR MONEY
| Metric | |||
|---|---|---|---|
| VIBE SCORE | 5.2 | 6.1 | 7.6 |
| VOICES | 28 | 187 | 14,071 |
| RANK | #397 | #161 | #2 |
| CATEGORY | Code | Code | Multimodal |
| OPEN/CLOSED | OPEN WEIGHTS | OPEN WEIGHTS | OPEN WEIGHTS |
| PRICE IN (per 1M) | $0.66 / 1M | $0.30 / 1M | $0.42 / 1M |
| PRICE OUT (per 1M) | $1.00 / 1M | $0.90 / 1M | $3.00 / 1M |
| THE CROWD LOVES | Not enough talk yet | Popular for coding help | Popular for coding help |
| THE CROWD GRUMBLES | Nobody is complaining yet | Nobody is complaining yet | Refuses or lectures more than people like |
ADD A MODEL (3 OF 4 SLOTS USED)
Fair fights only: chat and coding models compare with each other.
