Token costs are given in credits per token.
Click here to buy credits.

Model	Context Limits	Price (credits)	Speed (toks/sec)
DansPE v1.3.0 24B Click for details online Mistral-Small | 24B Creative, Roleplay, Assistant	32768 8192	0.200 0.800	12157.688323220365 30.713701251192955
DeepSeek V4 Flash Click for details online DeepSeekV4 | 284B Assistant, Tools, Coding, Smart	1048576 1048576	0.076 0.200	1904.8553788329673 31.66177939315591
DeepSeek V4 Flash 0731 Click for details online DeepSeekV4 | 284B Assistant, Tools, Coding, Smart	1048576 1048576	0.080 0.240	3904.6177732130514 64.58545042894325
GLM 4.7 Click for details online GLM MoE | 353B Creative, Roleplay, Assistant, Tools, Coding, Smart	131072 131072	0.280 1.000	399.11425826656597 14.352733088051183
GPT OSS 120B Click for details online GPT | 120B Assistant, Tools, Coding, Smart	131072 131072	0.020 0.100	3944.9515556226584 43.03055061940143
Magnum 72B v4 Click for details online Qwen 2.5 | 72B Creative, Roleplay, Assistant, Coding, Smart	32768 4096	1.000 2.000	3244.689879905629 17.47830935553565
MythoLite Click for details online LLaMA 2 | 13B Creative, Roleplay	2560 150	FREE FREE	181.90773667479118 64.79899954752567
MythoMax Click for details online LLaMA 2 | 13B Creative, Roleplay	8192 8192	0.140 0.240	7908.796901136739 69.20512934542487
Qwen3.8 27B Click for details online Qwen3 | 27B Assistant, Tools, Coding, Creative, Smart	262144 262144	0.080 1.000	1599.388391533339 64.65179986183439
ReMM-SLerp Click for details online LLaMA 2 | 13B Creative, Roleplay	6144 6144	0.140 0.260	7661.644082003876 49.66041967914925
Weaver Alpha Click for details online LLaMA 2 | 13B System Creative, Verbose	8000 6000	0.160 0.300	1176.6423897371437 61.196395580417565

DansPE v1.3.0 24B • DeepSeek V4 Flash • DeepSeek V4 Flash 0731 • GLM 4.7 • GPT OSS 120B • Magnum 72B v4 • MythoLite • MythoMax • Qwen3.8 27B • ReMM-SLerp • Weaver Alpha

The LLaMA v2 fine-tune that everyone's been raving about, extended to 8k for your prompting pleasure. Utterly unfiltered, rich descriptions, and unexpectedly good output.
Use the Alpaca instruct-prompt for best effect. Parameters should focus on Top-P instead of Top-K. Sometimes needs a high Rep.Pen. to get kicked out of a rut.
Click for model source.