GLM 5.3, uncensored
Yes. Both GLM 5.3 models are here without refusals: the full model as glm-5.3-xploded and the fast one as glm-5.3-flash-xploded.
| Model and id | Context | Input / 1M | Cached / 1M | Output / 1M |
|---|---|---|---|---|
| GLM 5.3glm-5.3-xploded | 1M | $2.00 | $1.00 | $5.00 |
| GLM 5.3 Flashglm-5.3-flash-xploded | 1M | $0.30 | $0.20 | $1.00 |
Full or Flash
They are closer than the price gap suggests. On the Artificial Analysis Intelligence Index the base GLM 5.3 scores 45 and base GLM 5.3 Flash scores 42, while Flash costs about a seventh as much for input and a fifth for output.
- Pick Flash for chat, roleplay, coding agents and anything with pictures. It reads images. The full model does not.
- Pick the full model when an answer has to be right the first time: long documents, careful analysis, multi step reasoning. It is text only.
Uncensored, not abliterated
These two are uncensored builds rather than abliterated ones: they were trained to answer, instead of having the refusal behaviour cut out afterwards. In use the difference is small. If you care about the distinction, this page explains it.
What it costs in practice
A 2,000 token prompt with a 600 token reply costs about $0.0012 on Flash and $0.0070 on the full model. Both take a million tokens of context and both call tools.
Calling it
curl https://uncensoredchats.com/v1/chat/completions \
-H "Authorization: Bearer $UNCENSORED_AI_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "glm-5.3-flash-xploded",
"messages": [{"role": "user", "content": "Hello"}]}'Keep reading
- Kimi K3.1 Pro, uncensored→
- DeepSeek V4.1 Flash, uncensored→
- Qwen3.8 27B, uncensored→
- The uncensored AI API→
- Set up SillyTavern→
- Set up opencode→
- Set up Claude Code→
- Abliterated vs uncensored→
- Cost calculator→
- Uncensored AI roleplay→
Updated 12 November 2026. Prices are read from the live price list.