PacketSafari

Customer-operated inference

Your cluster. Your boundary.

Run PacketSafari with an approved private-AI endpoint while the Core Engine keeps packet truth bounded, inspectable, and inside your deployment policy.

Design your deployment Explore cluster fit

8×H100 and 8×H200 are supported infrastructure profiles. Final model fit is confirmed for each deployment and workload.

Customer deployment boundaryEgress policy · enforced
01 / Packet storageCapture stays privateRetention · access · audit
bounded decode
02 / PacketSafariCore EngineFrames · flows · fields
compact evidence
03 / Qualified routeYour AI endpointPrivate · hosted · approved API
Codex remains the investigation engine
Model + infrastructure sizing

Private-AI deployment planner.

Already own hardware? See which supported models fit. Targeting a model? Compare the reference infrastructure it requires.

Compare capacity fit, PacketSafari support status, context, concurrency headroom, and indicative infrastructure cost. Hardware prices are indicative USD purchase ranges and exclude networking, facilities, support, and PacketSafari. Fast identifies a tested route that met the focused-case target; it is not a universal latency guarantee. Final runtime, topology, and investigation quality are confirmed during deployment qualification.

Supported infrastructure profileModel fit confirmed during deployment

8× H100 · 640 GB

A supported PacketSafari enterprise infrastructure profile for customer-operated AI, with room for larger models and concurrent investigations.

Indicative hardware price$250K–$400K
Supported and evaluated model fitQualification included
  • Kimi K2256K context
    Supported · Fast

    Recommended alternative model for a PacketSafari deployment.

  • Kimi K2.6Model-dependent context
    Supported · Fast

    Supported alternative model for a PacketSafari deployment.

  • MiniMax M3MXFP8 capacity fit context
    Not supported

    Capacity fit only; the model failed PacketSafari’s investigation-quality test.

  • DeepSeek V4 Flash 0731Quantized capacity fit context
    Supported · Fast

    Supported focused-model option; exact runtime fit is confirmed during deployment.

  • Qwen3 235BRevision-dependent context
    Not supported

    Capacity fit only.

  • Qwen3 Coder (large)Revision-dependent context
    Not supported

    Capacity fit only.

Frontier deployment comparison

Choose where frontier intelligence runs.

Compare a managed frontier route with supported and candidate private-AI options. The deployment boundary changes; PacketSafari’s packet evidence contract does not.

01Managed frontier

OpenAI GPT-5.6 Sol

Position
Hosted strongest-model route
AI boundary
Approved managed provider
Infrastructure
No customer GPU cluster
Published context
1.05M tokens
Availability
SaaS and approved hosted deployments
02Supported private reference

Kimi K2

Position
Recommended on-prem starting point
AI boundary
Customer-operated endpoint
Infrastructure
8× H100 or 8× H200 reference profiles
Published context
256K tokens
Availability
After deployment qualification
03Private frontier evaluation

Kimi K3

Position
High-capacity on-prem evaluation
AI boundary
Customer-operated endpoint
Infrastructure
8× B300 or 8× AMD MI355X reference profiles
Published context
1M tokens
Availability
Not currently supported
What the number meansContext is the model’s working envelope—not the PCAP size limit.

PacketSafari keeps packet processing in the Core Engine and supplies compact, relevant evidence to the Agent. The usable context configured for a deployment may be lower than the model’s published maximum.

Shared foundationPacketSafari Core Engine

Bounded packet evidence · inspectable findings · one investigation workflow

Current model evidence

Supported means tested. Not merely listed.

Finding a model through an endpoint is only discovery. PacketSafari separates API compatibility, investigation quality, and the hardware profile that will carry the workload.

01
Supported · Fast

Kimi K2

256K tokens context
Recommended qualified option

The recommended alternative-model starting point for customer-operated inference.

02
Supported

GLM 5.2

1M tokens context
Qualified deployment option

Available as an alternative model for qualified customer-operated deployments.

03
Supported · Fast

DeepSeek V4 Flash 0731

Model-dependent context
Supported

A tested focused-model option; exact endpoint compatibility is confirmed during onboarding.

04
Experimental

Kimi K3

1M tokens context
Not currently supported

Available only for evaluation and future deployment assessment.

05
Not supported

Qwen3 Coder 30B

256K native · up to 1M extended context
Not currently supported

Available only for evaluation and future deployment assessment.

06
Not supported

Qwen3 235B

Revision-dependent context
Not currently supported

Available only for evaluation and future deployment assessment.

One qualification boundary

Qualify once. Monitor drift without spending.

A short validation uses synthetic data only. Ongoing monitoring checks for material changes without touching a customer capture.

  1. 01Discover

    Confirm your selected model and deployment profile.

  2. 02Qualify

    Validate compatibility without using a customer PCAP.

  3. 03Accept

    Confirm investigation quality and operating capacity.

  4. 04Monitor

    Keep the approved deployment profile current.

Validated for your deployment. PacketSafari confirms the selected model, endpoint, workload, and capacity profile before production use.

Private AI deployment

Start with your captures, cluster, and security boundary.

Plan an on-prem validation Read the on-prem AI docs