Prices per 1M tokens
Uncensored coding model prices per 1M tokens
Uncensored coding models cost $0.24 to $3.00 per 1M input tokens and $0.90 to $8.00 per 1M output tokens at Silk Compute, read November 18, 2026.
bestuncensoredaiforcoding.com is published by the team behind Silk Compute, one of the providers ranked here.
Every hosted model, by API
| Model | API | Input / 1M | Output / 1M | Note |
|---|---|---|---|---|
| Kimi K3.1 Pro | Silk ComputeKimi K3.1 Pro | $3.00 | $8.00 | |
| DeepSeek V4.1 Flash | Silk ComputeDeepSeek V4.1 Flash 1017 | $0.24 | $0.90 | |
| DeepSeek V4.1 Flash | MorphDeepSeek V4.1 Flash Uncensored | $0.30 | $1.20 | Approval required |
| DeepSeek V4.1 Flash | abliteration.aiAbliterated Large Flash | $1.00 | $2.00 | |
| GLM-5.3 | Silk ComputeGLM-5.3 | $2.00 | $4.00 | |
| GLM-5.3 | abliteration.aiAbliterated Large v2 (GLM-5.3) | $3.00 | $5.00 | |
| GLM-5.3 Flash | Silk ComputeGLM-5.3 Flash | $0.24 | $1.00 |
From each ranked provider's pricing page, read November 18, 2026.
Silk Compute, with cached input
Agents resend the conversation every turn, so most input bills at the cached rate.
| Model | Input | Cached input | Output | Max output tokens |
|---|---|---|---|---|
| Kimi K3.1 Pro | $3.00 | $0.30 | $8.00 | 943,718 |
| DeepSeek V4.1 Flash | $0.24 | $0.038 | $0.90 | 131,072 |
| GLM-5.3 | $2.00 | $0.30 | $4.00 | 131,072 |
| GLM-5.3 Flash | $0.24 | $0.06 | $1.00 | 131,072 |
USD per 1M tokens, from silkcompute.com/pricing.
Plans and credits
Each monthly plan includes more usage than it costs, and prepaid credits start at $5 and never expire.
| Plan | Monthly | Yearly | Usage per month |
|---|---|---|---|
| Go | $10 | $100 | $12 |
| Pro | $20 | $200 | $25 |
| Max | $50 | $500 | $70 |
| Ultra | $200 | $2000 | $300 |
abliteration.ai's plans of $20, $50 and $200 a month each include usage equal to the price.
Each API's main coding model
The API table lists each one with its price per 1M tokens.
What decides the bill
- Keep Kimi K3.1 Pro for the hard part and send subagents to DeepSeek V4.1 Flash or GLM-5.3 Flash.
- Ask for diffs instead of whole files, since output costs about two to four times input.