Best Uncensored LLMsUncensored DeepSeek V4.1 Flash vs GLM 5.3: which one to use
The two most popular uncensored model families, head to head on refusals, agent tasks, speed and price, with every uncensored version of each and where to run them.
Uncensored builds of DeepSeek V4.1 Flash and Z.ai's GLM 5.3 are the two families people ask about most. Both have several versions from different makers. Here are the best of each, side by side.
Short answerThe best of each are effectively tied (69.8 vs 69.7). GLM finishes more agent tasks; The GLM build is cheaper.
| Best of each | DeepSeek V4.1 Flash 1017 Uncensored | GLM 5.3 Flash Uncensored |
|---|---|---|
| Score | 69.8 | 69.7 |
| Answers edgy requests | 100% | 94% |
| Agent tasks finished | 2 of 5 | 3 of 5 |
| Loops | 0% | 40% |
| Tokens per second | 278 | 25 |
| $ per 1M in + out | $1.65 | $1.00 |
| Context | 1M | 1M |
Every uncensored DeepSeek V4.1 Flash
| Rank | Model | Score | Agent tasks | Loops | Answers edgy | Hard-line refusals | $ / 1M in+out | Where to run it |
|---|---|---|---|---|---|---|---|---|
| 1 | DeepSeek V4.1 Flash 1017 Uncensored | 69.8 | 2/5 | 0% | 100% | 0% | $1.65 | VeniceNanoGPTOpenUncensored |
| 4 | DeepSeek V4.1 Flash 0910 Uncensored | 66.1 | 2/5 | 20% | 97% | 0% | $1.65 | NanoGPTOpenUncensored |
| 5 | DeepSeek V4.1 Flash Uncensored | 64.3 | 1/5 | 0% | 100% | 0% | $1.65 | VeniceNanoGPTOpenUncensored |
Every uncensored GLM 5.3
| Rank | Model | Score | Agent tasks | Loops | Answers edgy | Hard-line refusals | $ / 1M in+out | Where to run it |
|---|---|---|---|---|---|---|---|---|
| 2 | GLM 5.3 Flash Uncensored | 69.7 | 3/5 | 40% | 94% | 14% | $1.00 | NanoGPTOpenUncensored |
| 8 | Apex Flash 1 Abliterated | 63.2 | 3/5 | 0% | 86% | 100% | $0.65 | NanoGPTOpenUncensored |
| 9 | GLM 5.3 Uncensored | 63.0 | 2/5 | 20% | 92% | 0% | $3.50 | NanoGPTOpenUncensored |