LOADING COMPARISON
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
HEAD TO HEAD
Qwen3.8 27B vs DeepSeek R1 vs Apertus v1.5 70B
Community vibe score against API price per 1M tokens. Up and left wins: loved and cheap.
WHAT PEOPLE SAY
Qwen3.8 27B
LOVE
Popular for coding help
COMPLAINT
Refuses or lectures more than people like
DeepSeek R1
LOVE
Popular for coding help
COMPLAINT
People find it expensive
VALUE FOR MONEY
NOT ON THE CHART (NO PUBLIC API PRICE):DeepSeek R1Apertus v1.5 70BSTILL IN THE TABLE BELOW. THE CHART ONLY PLOTS MODELS WITH A KNOWN PRICE PER 1M TOKENS.
| Metric | |||
|---|---|---|---|
| VIBE SCORE | 7.6 | 6.6 | 7.4 |
| VOICES | 16,128 | 2,906 | 2,527 |
| RANK | #2 | #66 | #13 |
| CATEGORY | Multimodal | Text | Text |
| OPEN/CLOSED | OPEN WEIGHTS | OPEN WEIGHTS | OPEN WEIGHTS |
| PRICE IN (per 1M) | $0.42 / 1M | N/A | N/A |
| PRICE OUT (per 1M) | $3.00 / 1M | N/A | N/A |
| THE CROWD LOVES | Popular for coding help | Popular for coding help | Not enough talk yet |
| THE CROWD GRUMBLES | Refuses or lectures more than people like | People find it expensive | Nobody is complaining yet |
ADD A MODEL (3 OF 4 SLOTS USED)
Fair fights only: chat and coding models compare with each other.
