Models for roleplay

Our roleplay picks are editorial, not a benchmark ranking.

4 models

Gemini 3.8 Flash50% off
Context
1M
Input / 1M
~~$1.575~~$0.788
Output / 1M
~~$7.875~~$3.938
Max output
66K
Accepts
- Text
- Image
- PDF
- Video
GLM 5.3 Flash
Context
1M
Input / 1M
$0.158
Output / 1M
$0.525
Max output
131K
Accepts
- Text
- Image
GPT 5.6 Luna
Context
1.1M
Input / 1M
$0.21
Output / 1M
$1.26
Max output
128K
Accepts
- Text
- Image
- PDF
DeepSeek V4 Flash 0731
Context
1M
Input / 1M
$0.0798
Output / 1M
$0.161
Max output
384K
Accepts
- Text

Sortable and filterable MiniRouter model catalog
					Accepts
Gemini 3.8 Flash50% off	1M	66K	~~$1.575~~$0.788	~~$7.875~~$3.938	Text Image PDF Video
GLM 5.3 Flash	1M	131K	$0.158	$0.525	Text Image
GPT 5.6 Luna	1.1M	128K	$0.21	$1.26	Text Image PDF
DeepSeek V4 Flash 0731	1M	384K	$0.0798	$0.161	Text

.model-list-enhanced { display: none !important; }JavaScript is off, so sorting and filtering are unavailable. The complete catalog is listed below.

Complete MiniRouter model catalog
Model	Context	Max output	Input / 1M	Output / 1M	Accepts	Providers
GLM 5.3 Flash`zai/glm-5.3-flash`	1M	131K	$0.158	$0.525	Text Image	vercel
GPT 5.6 Luna`openai/gpt-5.6-luna`	1.1M	128K	$0.21	$1.26	Text Image PDF	vercel
DeepSeek V4 Flash 0731`deepseek/deepseek-v4-flash-0731`	1M	384K	$0.0798	$0.161	Text	vercel
Gemini 3.8 Flash50% off`google/gemini-3.8-flash`	1M	66K	~~$1.575~~$0.788	~~$7.875~~$3.938	Text Image PDF Video	vercel

Reliability notes

Some of these have nowhere to fail over to. GLM 5.3 Flash, GPT 5.6 Luna, DeepSeek V4 Flash 0731, Gemini 3.8 Flash are served by a single registered upstream. If it goes offline, these models are unavailable until it recovers. How failover works.

Client setup for SillyTavern, RisuAI and Agnai — and what we store, which is not your prompts — is on /use/roleplay. The arithmetic for what one message costs at these rates is in the cost guide.