Heabsy

GLB

15 models available

Heabsy states that its EEA Qwen3.8 deployment processes prompts and completions in memory without durable storage, logging, or training use. Data handling and deployment region vary by model; see each model’s provider details. Its GDPR Art. 28 Data Processing Addendum is available at https://heabsy.com/dpa.

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 Heabsy

15 models

Model	Context	Input /M	Output /M	Cache read /M	Cache hit	TTFT	Throughput
CyberHeabsy NVFP4	1.0M	$1.50/M	$4.50/M	$0.27/M	76.6%	2.2s	127.5 tps
DeepSeek V4.1 Flash Uncensored FP8	1.0M	$0.42/M	$1.47/M	$0.042/M	93.4%	2.8s	443.4 tps
DeepSeek V4.1 Flash Uncensored Thinking FP8	1.0M	$0.42/M	$1.47/M	$0.042/M	72.4%	0.7s	280 tps
GLM 5.3 Flash Uncensored NVFP4	1.0M	$0.20/M	$0.80/M	$0.0641/M	90.6%	1.8s	163.7 tps
GLM 5.3 Uncensored NVFP4	1.0M	$1.65/M	$2.98/M	$0.397/M	N/A	N/A	N/A
MiMo V2.6 Flash NVFP4	1.0M	$0.50/M	$1.50/M	$0.20/M	N/A	N/A	N/A
MiMo V2.6 Flash Abliterated NVFP4	1.0M	$0.50/M	$1.50/M	$0.27/M	86.7%	6.4s	173 tps
MiMo V2.6 Flash Uncensored NVFP4	1.0M	$0.462/M	$1.50/M	$0.147/M	97%	1s	205.9 tps
Qwen 3.8 27B Fable NVFP4	524k	$0.0998/M	$0.609/M	$0.047/M	0%	3.4s	124.1 tps
Qwen 3.8 27B Queen NVFP4	524k	$0.0998/M	$0.609/M	$0.047/M	37.8%	14.3s	216.1 tps
Qwen 3.8 27B Uncensored FP8	524k	$0.437/M	$1.97/M	$0.0569/M	84.6%	1.4s	270.5 tps
Qwen 3.8 27B Uncensored Fast by Heabsy NVFP4	524k	$0.83/M	$2.62/M	$0.15/M	8.6%	0.6s	186.2 tps
Qwen 3.8 27B Uncensored Thinking FP8	524k	$0.437/M	$1.97/M	$0.0569/M	37.7%	1.9s	225.9 tps
Qwen3.8 27B NVFP4	262k	$1.05/M	$3.15/M	$0.0799/M	N/A	0.9s	176 tps
Qwen3.8 27B Thinking NVFP4	262k	$1.05/M	$3.15/M	$0.0799/M	N/A	0.9s	176 tps

CyberHeabsy

NVFP4

Context: 1.0M
Input /M: $1.50/M
Output /M: $4.50/M
Cache read /M: $0.27/M
Cache hit: 76.6%
TTFT: 2.2s
Throughput: 127.5 tps

DeepSeek V4.1 Flash Uncensored

FP8

Context: 1.0M
Input /M: $0.42/M
Output /M: $1.47/M
Cache read /M: $0.042/M
Cache hit: 93.4%
TTFT: 2.8s
Throughput: 443.4 tps

DeepSeek V4.1 Flash Uncensored Thinking

FP8

Context: 1.0M
Input /M: $0.42/M
Output /M: $1.47/M
Cache read /M: $0.042/M
Cache hit: 72.4%
TTFT: 0.7s
Throughput: 280 tps

GLM 5.3 Flash Uncensored

NVFP4

Context: 1.0M
Input /M: $0.20/M
Output /M: $0.80/M
Cache read /M: $0.0641/M
Cache hit: 90.6%
TTFT: 1.8s
Throughput: 163.7 tps

GLM 5.3 Uncensored

NVFP4

Context: 1.0M
Input /M: $1.65/M
Output /M: $2.98/M
Cache read /M: $0.397/M
Cache hit: N/A
TTFT: N/A
Throughput: N/A

MiMo V2.6 Flash

NVFP4

Context: 1.0M
Input /M: $0.50/M
Output /M: $1.50/M
Cache read /M: $0.20/M
Cache hit: N/A
TTFT: N/A
Throughput: N/A

MiMo V2.6 Flash Abliterated

NVFP4

Context: 1.0M
Input /M: $0.50/M
Output /M: $1.50/M
Cache read /M: $0.27/M
Cache hit: 86.7%
TTFT: 6.4s
Throughput: 173 tps

MiMo V2.6 Flash Uncensored

NVFP4

Context: 1.0M
Input /M: $0.462/M
Output /M: $1.50/M
Cache read /M: $0.147/M
Cache hit: 97%
TTFT: 1s
Throughput: 205.9 tps

Qwen 3.8 27B Fable

NVFP4

Context: 524k
Input /M: $0.0998/M
Output /M: $0.609/M
Cache read /M: $0.047/M
Cache hit: 0%
TTFT: 3.4s
Throughput: 124.1 tps

Qwen 3.8 27B Queen

NVFP4

Context: 524k
Input /M: $0.0998/M
Output /M: $0.609/M
Cache read /M: $0.047/M
Cache hit: 37.8%
TTFT: 14.3s
Throughput: 216.1 tps

Qwen 3.8 27B Uncensored

FP8

Context: 524k
Input /M: $0.437/M
Output /M: $1.97/M
Cache read /M: $0.0569/M
Cache hit: 84.6%
TTFT: 1.4s
Throughput: 270.5 tps

Qwen 3.8 27B Uncensored Fast by Heabsy

NVFP4

Context: 524k
Input /M: $0.83/M
Output /M: $2.62/M
Cache read /M: $0.15/M
Cache hit: 8.6%
TTFT: 0.6s
Throughput: 186.2 tps

Qwen 3.8 27B Uncensored Thinking

FP8

Context: 524k
Input /M: $0.437/M
Output /M: $1.97/M
Cache read /M: $0.0569/M
Cache hit: 37.7%
TTFT: 1.9s
Throughput: 225.9 tps

Qwen3.8 27B

NVFP4

Context: 262k
Input /M: $1.05/M
Output /M: $3.15/M
Cache read /M: $0.0799/M
Cache hit: N/A
TTFT: 0.9s
Throughput: 176 tps

Qwen3.8 27B Thinking

NVFP4

Context: 262k
Input /M: $1.05/M
Output /M: $3.15/M
Cache read /M: $0.0799/M
Cache hit: N/A
TTFT: 0.9s
Throughput: 176 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 Heabsy via the API

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

cURL

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

For models with provider selection, append :heabsy to the model ID, set "provider": "heabsy" in the request body, or send an X-Provider: heabsy 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.

Heabsy terms of use Heabsy privacy policy

Model

Context

Input /M

Output /M

Cache read /M

Cache hit

TTFT

Throughput

CyberHeabsy

NVFP4

1.0M

$1.50/M

$4.50/M

$0.27/M

76.6%

2.2s

127.5 tps

DeepSeek V4.1 Flash Uncensored

FP8

1.0M

$0.42/M

$1.47/M

$0.042/M

93.4%

2.8s

443.4 tps

DeepSeek V4.1 Flash Uncensored Thinking

FP8

1.0M

$0.42/M

$1.47/M

$0.042/M

72.4%

0.7s

280 tps

GLM 5.3 Flash Uncensored

NVFP4

1.0M

$0.20/M

$0.80/M

$0.0641/M

90.6%

1.8s

163.7 tps

GLM 5.3 Uncensored

NVFP4

1.0M

$1.65/M

$2.98/M

$0.397/M

N/A

MiMo V2.6 Flash

NVFP4

1.0M

$0.50/M

$1.50/M

$0.20/M

N/A

MiMo V2.6 Flash Abliterated

NVFP4

1.0M

$0.50/M

$1.50/M

$0.27/M

86.7%

6.4s

173 tps

MiMo V2.6 Flash Uncensored

NVFP4

1.0M

$0.462/M

$1.50/M

$0.147/M

97%

205.9 tps

Qwen 3.8 27B Fable

NVFP4

524k

$0.0998/M

$0.609/M

$0.047/M

3.4s

124.1 tps

Qwen 3.8 27B Queen

NVFP4

524k

$0.0998/M

$0.609/M

$0.047/M

37.8%

14.3s

216.1 tps

Qwen 3.8 27B Uncensored

FP8

524k

$0.437/M

$1.97/M

$0.0569/M

84.6%

1.4s

270.5 tps

Qwen 3.8 27B Uncensored Fast by Heabsy

NVFP4

524k

$0.83/M

$2.62/M

$0.15/M

8.6%

0.6s

186.2 tps

Qwen 3.8 27B Uncensored Thinking

FP8

524k

$0.437/M

$1.97/M

$0.0569/M

37.7%

1.9s

225.9 tps

Qwen3.8 27B

NVFP4

262k

$1.05/M

$3.15/M

$0.0799/M

N/A

0.9s

176 tps

Qwen3.8 27B Thinking

NVFP4

262k

$1.05/M

$3.15/M

$0.0799/M

N/A

0.9s

176 tps

Using Heabsy via the API

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

cURL

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

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