All comparisons
Kimi K3.1 Pro vs GLM-5.3
Abliterated against abliterated, 2,780B against 753B.
DeepSeek V4.1 Flash 1017 vs GLM-5.3 Flash
Abliterated against abliterated, 763B against 320B.
Kimi K3.1 Pro vs Hermes 3 405B
Abliterated against fine-tuned, 2,780B against 405B.
Qwen 3.8 27B Abliterated vs DeepSeek V4.1 Flash 1017
Abliterated against abliterated, 27.8B against 763B.
Silk Compute vs Abliteration
Two unrestricted AI APIs, point by point.
Silk Compute vs Morph
Two unrestricted AI APIs, point by point.
All models at once
| Kimi K3.1 Pro | DeepSeek V4.1 Flash 1017 | GLM-5.3 | GLM-5.3 Flash | Hermes 3 405B | |
|---|---|---|---|---|---|
| Refusals | |||||
| How refusals were removed | Abliterated | Abliterated | Abliterated | Abliterated | Fine-tuned |
| Size and context | |||||
| Parameterstotal | 2,780B | 763B | 753B | 320B | 405B |
| Active parametersper token | All | All | All | 18B | All |
| Context windowtokens | 1,048,576 | 1,048,576 | 1,048,576 | 1,048,576 | 131,072 |
| Max outputtokens | 943,718 | 131,072 | 131,072 | 131,072 | Not published |
| Inputs | |||||
| Text | |||||
| Images | |||||
| Video | |||||
| Price, lowest listed host | |||||
| InputUSD per 1M | $3.50 | $0.24 | $2.00 | $0.24 | $1.20 |
| OutputUSD per 1M | $8.00 | $0.90 | $4.00 | $1.00 | $1.20 |
| BlendedUSD per 1M | $4.63 | $0.41 | $2.50 | $0.43 | $1.20 |
| Where it runs | |||||
| Hosts | Silk Compute | Silk Compute, Morph, Abliteration | Silk Compute, Abliteration | Silk Compute | Unfil.AI |
| API formatsat the host | 3 of 3 | 3 of 3 | 3 of 3 | 3 of 3 | 1 of 3 |
| Runs locally | |||||