Skip to content

blackbit.void — ai for security teams

voidzerosoon
get api key
void v1 is publicly available — try it now

the ai gateway for security teams

Meet VOID — an AI gateway for authorized security testing, code review, and development.

get api key

built by engineers previously at

  • Lockheed Martin
  • Palantir
  • Crypto.com

platform capabilities

01

ai for security work

Use AI models for authorized security testing, code review, and development through one gateway.

02

pay as you go

Add prepaid credits and use them as you go. No subscription required.

03

zero inference retention

Prompts and outputs are not retained after inference. Cloud sandbox data is deleted after each task; billing records are separate.

04

confidential computing

Inference runs in protected compute environments with encryption designed to protect customer content during processing.

01 — model catalog

23 public models

model
input
output
DeepSeek V4.1 Flash
input$0.15
output$0.6
GLM 5.3
input$0.7
output$2.2
Qwen 3.8 Max
input$2
output$6
DeepSeek V4 Flash 0731
input$0.06
output$0.153
Laguna S 2.1
input$0.1
output$0.2

showing 05 of 23 models

02 — benchmark

23 models — lower is better

cost per intelligence

Weighted USD per task, split by token type. Lower is better.

  • answer
  • reasoning
  • cache write
  • cache hit
  • input

  1. DeepSeek V4 Flash 0731

    $0.04

    Cost breakdown: answer: $0.021; reasoning: $0.009; cache write: $0.006; cache hit: $0.000; input: $0.005.

  2. Laguna S 2.1

    $0.06

    Cost breakdown: answer: $0.028; reasoning: $0.012; cache write: $0.010; cache hit: $0.001; input: $0.008.

  3. DeepSeek V4 Flash

    $0.08

    Cost breakdown: answer: $0.036; reasoning: $0.016; cache write: $0.013; cache hit: $0.001; input: $0.010.

  4. DeepSeek V4.1 Flash

    $0.15

    Cost breakdown: answer: $0.084; reasoning: $0.036; cache write: $0.015; cache hit: $0.001; input: $0.012.

  5. MiniMax M3

    $0.30

    Cost breakdown: answer: $0.168; reasoning: $0.072; cache write: $0.030; cache hit: $0.002; input: $0.024.

  6. Mistral Large 3

    $0.39

    Cost breakdown: answer: $0.210; reasoning: $0.090; cache write: $0.050; cache hit: $0.004; input: $0.040.

  7. GLM 5.3

    $0.57

    Cost breakdown: answer: $0.308; reasoning: $0.132; cache write: $0.070; cache hit: $0.006; input: $0.056.

  8. GLM 5.2

    $0.66

    Cost breakdown: answer: $0.357; reasoning: $0.153; cache write: $0.080; cache hit: $0.006; input: $0.064.

  9. GLM 5

    $0.83

    Cost breakdown: answer: $0.448; reasoning: $0.192; cache write: $0.100; cache hit: $0.008; input: $0.080.

  10. Kimi K2.7 Code

    $0.98

    Cost breakdown: answer: $0.560; reasoning: $0.240; cache write: $0.095; cache hit: $0.008; input: $0.076.

  11. Kimi K2.6

    $0.98

    Cost breakdown: answer: $0.560; reasoning: $0.240; cache write: $0.095; cache hit: $0.008; input: $0.076.

  12. DeepSeek V4 Pro

    $1.02

    Cost breakdown: answer: $0.487; reasoning: $0.209; cache write: $0.174; cache hit: $0.014; input: $0.139.

  13. GLM 5 Turbo

    $1.03

    Cost breakdown: answer: $0.560; reasoning: $0.240; cache write: $0.120; cache hit: $0.010; input: $0.096.

  14. GLM 5.1

    $1.14

    Cost breakdown: answer: $0.616; reasoning: $0.264; cache write: $0.140; cache hit: $0.011; input: $0.112.

  15. Abliterated Model

    $1.16

    Cost breakdown: answer: $0.420; reasoning: $0.180; cache write: $0.300; cache hit: $0.024; input: $0.240.

  16. Qwen 3.8 Max

    $1.58

    Cost breakdown: answer: $0.840; reasoning: $0.360; cache write: $0.200; cache hit: $0.016; input: $0.160.

  17. Abliterated Model Large

    $1.94

    Cost breakdown: answer: $0.700; reasoning: $0.300; cache write: $0.500; cache hit: $0.040; input: $0.400.

  18. Abliterated Model Large v2

    $1.94

    Cost breakdown: answer: $0.700; reasoning: $0.300; cache write: $0.500; cache hit: $0.040; input: $0.400.

  19. Qwen 3.7 Max

    $1.97

    Cost breakdown: answer: $1.050; reasoning: $0.450; cache write: $0.250; cache hit: $0.020; input: $0.200.

  20. Kimi K3

    $3.56

    Cost breakdown: answer: $2.100; reasoning: $0.900; cache write: $0.300; cache hit: $0.024; input: $0.240.

  21. CyberGLM

    $3.75

    Cost breakdown: answer: $2.100; reasoning: $0.900; cache write: $0.400; cache hit: $0.032; input: $0.320.

  22. STEALTH-MODEL-01

    $6.94

    Cost breakdown: answer: $4.200; reasoning: $1.800; cache write: $0.500; cache hit: $0.040; input: $0.400.

  23. CyberKimi

    $7.50

    Cost breakdown: answer: $4.200; reasoning: $1.800; cache write: $0.800; cache hit: $0.064; input: $0.640.
Sample figures from a fixed task mix at published Console token rates. Flash models stay cheap; max, pro, and cyber routes cost more.

AI should help teams
understand their systems,
improve their security, and
better defend themselves
with clarity and control.

get api key

03 — use cases

04 workflows — security teams

uc-01use case

security testing

Use AI to help identify weaknesses in systems you own or have written permission to test. Review findings before making changes.

scope

approved

focus

testing

output

findings

uc-02use case

code review

Review code for potential security issues and explore fixes before deployment. Validate suggestions in your own development workflow.

scope

code

focus

review

output

changes

uc-03use case

security reporting

Use findings to support internal security and compliance reviews. Save the reports you need; Blackbit does not keep a findings archive.

scope

internal

focus

review

output

reports

uc-04use case

confidential workflows

Work with sensitive code and task context using confidential inference. Run agents locally or in cloud sandboxes with task-limited storage.

scope

content

focus

privacy

output

analysis

Give your team
the tools to
{defend themselves}

get api key