Kimi K3.1 Pro
Kimi K3.1 Pro is the model for large jobs, such as multi-file changes, long refactors and authorized security work.
The best uncensored AI for coding is Kimi K3.1 Pro, an abliterated 2,780B-parameter model with a 1M-token context that runs in Claude Code, Codex, OpenCode and Cursor.
Four hosted models, then four you run yourself.
How we rank| # | Model | Parameters | Context | Max output | Price per 1M, in / out | Inputs | Pick |
|---|---|---|---|---|---|---|---|
| Hosted, through an API | |||||||
| 1 | Kimi K3.1 ProMoonshot AI · abliterated | 2,780B | 1M | 943,718 | $3.00 / $8.00 | Text, Images, Video | Best overall |
| 2 | DeepSeek V4.1 Flash 1017DeepSeek · abliterated | 763B | 1M | 131,072 | $0.24 / $0.90 | Text, Images | Best value |
| 3 | GLM-5.3Z.ai · abliterated | 753B | 1M | 131,072 | $2.00 / $4.00 | Text | Best text-only coder |
| 4 | GLM-5.3 FlashZ.ai · abliterated | 320B total, 18B active | 1M | 131,072 | $0.24 / $1.00 | Text, Images | Best for volume |
| Local, on a GPU you own | |||||||
| 5 | Qwen 3.8 27B Abliteratedhuihui-ai · runs locally | 27.8B | Your memory | Your server | about 17 GB at 4-bit | Text | Best local overall |
| 6 | Qwen 3.6 35B-A3B Abliteratedhuihui-ai · runs locally | 36B total, 3B active | Your memory | Your server | about 22 GB at 4-bit | Text | Best local mixture-of-experts |
| 7 | Gemma 4 31B Hereticcoder3101 · runs locally | 31.3B | Your memory | Your server | about 19 GB at 4-bit | Text | Best local Gemma |
| 8 | Llama 3.1 8B Abliteratedmlabonne · runs locally | 8B | Your memory | Your server | about 5 GB at 4-bit | Text | Best for a laptop |
Hosted prices are Silk Compute's, read November 18, 2026, and 4-bit sizes are estimates.
Kimi K3.1 Pro is the model for large jobs, such as multi-file changes, long refactors and authorized security work.
DeepSeek V4.1 Flash 1017 is the default for most coding sessions, because cached input at $0.038 per 1M tokens keeps long agent sessions cheap.
GLM-5.3 is a text-only coder from another model family, worth asking for a second opinion on a hard change.
GLM-5.3 Flash is the model for volume, such as bulk edits, test generation and many subagents at once.
The newest abliterated Qwen here, small enough at 4-bit to leave room for context on a 24 GB card.
The only mixture-of-experts model among the local picks.
Google's Gemma 4 31B, edited to stay close to the original model.
The smallest and oldest model here, so expect much less from it in an agent than from the 27B to 36B models.
The API format each agent speaks decides it.
Full matrixEach config adds every hosted model and starts on DeepSeek V4.1 Flash 1017.
Step-by-step guides{
"apiKeyHelper": "echo $SILK_API_KEY",
"env": {
"ANTHROPIC_BASE_URL": "https://api.silkcompute.com",
"ANTHROPIC_DEFAULT_MODEL": "deepseek-v4.1-flash-1017-uncensored",
"ANTHROPIC_DEFAULT_FABLE_MODEL": "kimi-k3.1-pro-uncensored",
"ANTHROPIC_DEFAULT_OPUS_MODEL": "kimi-k3.1-pro-uncensored",
"ANTHROPIC_DEFAULT_SONNET_MODEL": "deepseek-v4.1-flash-1017-uncensored",
"ANTHROPIC_DEFAULT_HAIKU_MODEL": "deepseek-v4.1-flash-1017-uncensored",
"CLAUDE_CODE_MAX_CONTEXT_TOKENS": "1048576"
},
"modelPicker": {
"replaceBuiltInOptions": true,
"options": [
{ "model": "kimi-k3.1-pro-uncensored", "label": "Kimi K3.1 Pro" },
{
"model": "deepseek-v4.1-flash-1017-uncensored",
"label": "DeepSeek V4.1 Flash 1017"
},
{
"model": "deepseek-v4.1-flash-1017-fast-uncensored",
"label": "DeepSeek V4.1 Flash 1017 Fast"
},
{ "model": "glm-5.3-uncensored", "label": "GLM-5.3" },
{ "model": "glm-5.3-flash-uncensored", "label": "GLM-5.3 Flash" },
{ "model": "hy4-uncensored", "label": "Hy4" }
]
}
}model = "deepseek-v4.1-flash-1017-uncensored"
model_provider = "silkcompute"
web_search = "disabled"
[model_providers.silkcompute]
name = "Silk Compute"
base_url = "https://api.silkcompute.com"
env_key = "SILK_API_KEY"
model_catalog_url = "https://api.silkcompute.com/codex/models"{
"$schema": "https://opencode.ai/config.json",
"model": "silkcompute/deepseek-v4.1-flash-1017-uncensored",
"provider": {
"silkcompute": {
"npm": "@ai-sdk/openai-compatible",
"name": "Silk Compute",
"options": {
"baseURL": "https://api.silkcompute.com",
"apiKey": "{env:SILK_API_KEY}"
},
"models": {
"kimi-k3.1-pro-uncensored": {
"name": "Kimi K3.1 Pro",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image", "video"], "output": ["text"] },
"limit": { "context": 1048576, "output": 943718 },
"cost": { "input": 3, "output": 8, "cache_read": 0.3 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"deepseek-v4.1-flash-1017-uncensored": {
"name": "DeepSeek V4.1 Flash 1017",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image"], "output": ["text"] },
"limit": { "context": 1048576, "output": 131072 },
"cost": { "input": 0.24, "output": 0.9, "cache_read": 0.038 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"deepseek-v4.1-flash-1017-fast-uncensored": {
"name": "DeepSeek V4.1 Flash 1017 Fast",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image"], "output": ["text"] },
"limit": { "context": 1048576, "output": 131072 },
"cost": { "input": 0.24, "output": 0.9, "cache_read": 0.05 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"glm-5.3-uncensored": {
"name": "GLM-5.3",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text"], "output": ["text"] },
"limit": { "context": 1048576, "output": 131072 },
"cost": { "input": 2, "output": 4, "cache_read": 0.3 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"glm-5.3-flash-uncensored": {
"name": "GLM-5.3 Flash",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image"], "output": ["text"] },
"limit": { "context": 1048576, "output": 131072 },
"cost": { "input": 0.24, "output": 1, "cache_read": 0.06 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"hy4-uncensored": {
"name": "Hy4",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image"], "output": ["text"] },
"limit": { "context": 1048576, "output": 65536 },
"cost": { "input": 1.6, "output": 3.2, "cache_read": 0.3 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
}
}
}
}
}The configs read your key from SILK_API_KEY and also add Hy4, which is not ranked here.
APIs for coding agents compares where to run them.
Yes. Claude Code reads ANTHROPIC_BASE_URL, so any API that serves an abliterated model in the Anthropic Messages format works, as the Claude Code setup page shows.
A little, but the size of the base model matters far more, as abliterated vs fine-tuned explains.
Legitimate work that hosted assistants often decline, such as security tooling for systems you are authorized to test, reverse engineering your own software and scrapers for public data. It is not a way around the law or anyone else's permission.