Mancer

Data handling can vary by model and deployment. The privacy badge above shows the more conservative of the provider-wide fallback and the policies for routes on this page; each model page shows the policy for that specific route.

Models served by Mancer

Model name

9 models

Model	Context	Input /M	Output /M	TTFT	Throughput
DeepSeek V4 Flash FP8 deepseek/deepseek-v4-flash:mancer	1.0M	$0.20/M	$0.53/M	0.8s	53 tps
DeepSeek V4 Flash (Thinking) FP8 deepseek/deepseek-v4-flash:thinking:mancer	1.0M	$0.20/M	$0.53/M	0.8s	53 tps
DeepSeek V4 Flash 0731 FP8 deepseek/deepseek-v4-flash-0731:mancer	1.0M	$0.21/M	$0.63/M	1.2s	38 tps
DeepSeek V4 Flash 0731 (Thinking) FP8 deepseek/deepseek-v4-flash-0731:thinking:mancer	1.0M	$0.21/M	$0.63/M	1.2s	38 tps
GLM 4.7 FP4 z-ai/glm-4.7:mancer	131k	$0.73/M	$2.63/M	1.9s	11 tps
GLM 4.7 (Thinking) FP4 z-ai/glm-4.7:thinking:mancer	131k	$0.73/M	$2.63/M	1.9s	11 tps
GPT OSS 120B FP8 openai/gpt-oss-120b:mancer	131k	$0.05/M	$0.26/M	0.7s	32 tps
Qwen3.8 27B FP8 qwen3.8-27b:mancer	262k	$0.21/M	$2.63/M	1.2s	52 tps
Qwen3.8 27B Thinking FP8 qwen3.8-27b:thinking:mancer	262k	$0.21/M	$2.63/M	1.2s	52 tps

DeepSeek V4 Flash

FP8

deepseek/deepseek-v4-flash:mancer

Context

1.0M

Input /M

$0.20/M

Output /M

$0.53/M

TTFT

0.8s

Throughput

53 tps

DeepSeek V4 Flash (Thinking)

FP8

deepseek/deepseek-v4-flash:thinking:mancer

Context

1.0M

Input /M

$0.20/M

Output /M

$0.53/M

TTFT

0.8s

Throughput

53 tps

DeepSeek V4 Flash 0731

FP8

deepseek/deepseek-v4-flash-0731:mancer

Context

1.0M

Input /M

$0.21/M

Output /M

$0.63/M

TTFT

1.2s

Throughput

38 tps

DeepSeek V4 Flash 0731 (Thinking)

FP8

deepseek/deepseek-v4-flash-0731:thinking:mancer

Context

1.0M

Input /M

$0.21/M

Output /M

$0.63/M

TTFT

1.2s

Throughput

38 tps

GLM 4.7

FP4

z-ai/glm-4.7:mancer

Context

131k

Input /M

$0.73/M

Output /M

$2.63/M

TTFT

1.9s

Throughput

11 tps

GLM 4.7 (Thinking)

FP4

z-ai/glm-4.7:thinking:mancer

Context

131k

Input /M

$0.73/M

Output /M

$2.63/M

TTFT

1.9s

Throughput

11 tps

GPT OSS 120B

FP8

openai/gpt-oss-120b:mancer

Context

131k

Input /M

$0.05/M

Output /M

$0.26/M

TTFT

0.7s

Throughput

32 tps

Qwen3.8 27B

FP8

qwen3.8-27b:mancer

Context

262k

Input /M

$0.21/M

Output /M

$2.63/M

TTFT

1.2s

Throughput

52 tps

Qwen3.8 27B Thinking

FP8

qwen3.8-27b:thinking:mancer

Context

262k

Input /M

$0.21/M

Output /M

$2.63/M

TTFT

1.2s

Throughput

52 tps

Prices are per million tokens. TTFT is the observed time to first token. Selectable routes include the applicable provider-selection markup; fixed routes show the standard model price. Availability and pricing refresh continuously; each model page shows the live provider comparison.

Using Mancer via the API

Call a model served by Mancer using its displayed model ID, including the provider suffix when selection is supported.

curl https://nano-gpt.com/api/v1/chat/completions \
  -H "Authorization: Bearer $NANOGPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-v4-flash:mancer",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

For models with provider selection, append :mancer to the model ID, set "provider": "mancer" in the request body, or send an X-Provider: mancer header. Explicit provider selection adds a route-specific markup over that provider's base price; the prices in the table above already include it. Models without provider selection use the displayed model ID without a provider suffix or selection surcharge.

See the API documentation for provider preferences, price-aware routing, and error behavior.