Team access you control
Invite teammates with defined roles, rotate API keys and review company activity. Invitation links expire and require the invited, verified email address.
LLMPods at a glance
LLMPods provides regional GPU infrastructure for open-model inference. Choose deployment capacity around your model, workload and data requirements.
- Operator
- Disket Teknoloji Limited Sirketi · Istanbul/Türkiye
- Current compute location
- Istanbul / Türkiye
- Model families
- Llama · Qwen · DeepSeek · Mistral
- Planned regions; not available today
- Frankfurt / Germany · Canada
Explore models, deployment options and pricing
A workspace for production inference
Manage team access, project limits and request-level usage on LLMPods-owned GPU infrastructure in Türkiye.
Project budgets and limits
Set monthly project budgets, request-rate and output-token limits. Requests reserve estimated cost before inference; final charges use verified token measurements.
Trace usage, keep content private
Filter requests by project, status or request ID. Inspect latency, tokens and cost, then export CSV. Request records contain metadata rather than prompt and response content.
Built around your market
Your AI runs closer to home.
Latency, data location and invoice currency are infrastructure decisions. Make them part of your deployment, not an afterthought.
Latency you can feel.
Shorter network paths can help a chat response start sooner. Türkiye-based compute keeps local workloads close; model size, queueing and prompt length still shape first-token time.
Measure first token, P50 and P95 from your application.
Know your KVKK exposure.
Know where prompts, outputs, logs and support data travel. Processing model workloads in Türkiye can help limit cross-border transfers, while lawful processing, retention and access controls remain essential.
Local hosting alone does not establish KVKK compliance.
Transparent FX on every payment.
LLMPods prices usage in USD. Card payments are charged in TRY at the Central Bank of the Republic of Türkiye (TCMB) USD forex selling rate current at the time of payment, so the TRY amount for the same workload moves with the exchange rate.
The rate and the TRY amount are shown before you pay and recorded with the payment; invoices are issued in TRY and state the rate used.
Türkiye today. More regions planned.
Compute is hosted in Türkiye today. Frankfurt and Canada are planned locations, not currently available deployment regions.
- Live
Istanbul / Türkiye
Current compute location · TR-IST-01
- Planned
Frankfurt / Germany
Planned · no launch date announced
- Planned
Canada
Planned · no launch date announced
Questions before your first deployment.
Where does my model run?
Current compute is in Türkiye. Frankfurt and Canada are planned; requests from other markets can be served from Türkiye without moving the GPUs.
Is local compute always faster?
A shorter network route can reduce transport delay, but first-token time also depends on the model, queue and input size. Compare the same model and prompt from your actual users' location.
Does Türkiye hosting automatically mean KVKK compliance?
No. You still need to assess processing grounds, notices, retention, access and any transfers through integrations or support. Regional compute is one part of that assessment.
Which exchange rate applies when I pay?
Prices are in USD. When you pay by card, the TRY amount is calculated with the TCMB USD forex selling rate current at that moment and fixed for that payment session. The TRY cost of the same usage therefore changes with the exchange rate; VAT is shown separately.
01/System
API in.
Tokens out.
Every request crosses the same seven layers. Each layer reports what it did, so the path is never a black box.
01
Application
REQUEST7F39A21
METHODPOST
02
LLMPods API
AUTHOK
SCHEMAOPENAI
03
Regional router
REGIONTR-IST
ROUTINGAUTO
04
Model scheduler
MODELQWEN
QUEUE0
05
Compute Pod
POD018
STATEACTIVE
06
GPU
CLASSPRO 6000
ALLOCRESERVED
07
Token stream
STATUSSTREAMING
FIRST TOKENDEMO 24ms
LAYER 03 Developer control
LAYER 02 System intelligence
LAYER 01 Physical compute
LAYER 03 Developer
Illustrative telemetry · demonstration data
02/Developer experience
Change the endpoint.
Keep your stack.
LLMPods speaks the OpenAI schema. Point your existing SDK at a new base URL and keep the code around it.
llmpods.com/panel/v1
1from openai import OpenAI
2import os
3
4client = OpenAI(
5 base_url="https://llmpods.com/panel/v1",ONE LINE MIGRATION →
6 api_key=os.environ["LLMPODS_API_KEY"],
7)
8
9stream = client.chat.completions.create(
10 model="llama-70b-instruct",
11 messages=[{"role": "user", "content": "Hello, Pod."}],
12 stream=True,
13)
14
15for chunk in stream:
16 print(chunk.choices[0].delta.content or "", end="")
- 01OpenAI-compatible schema
- 02Streaming responses
- 03Tool calling
- 04Embeddings
- 05Familiar SDK support
- 06Developer-first migration
03/Pods
A Pod for
every workload.
Four deployment topologies on one runtime. The difference between them is how much of the machine is yours.
Topology 01
Shared Pod
For elastic inference.
- Usage-based
- Autoscaling
- Managed infrastructure
- Shared GPU pool
Topology 02
Performance Pod
For predictable production workloads.
- Reserved capacity
- Priority scheduling
- Consistent performance
- Predictable availability
Topology 03
Dedicated Pod
For isolated workloads.
- Dedicated GPU
- Custom models
- Private endpoint
- Isolated resources
Topology 04
Enterprise Pod
For strategic infrastructure requirements.
- Multi-GPU
- Dedicated cluster
- Private networking
- Regional control
- SLA
- Custom architecture
04/Scale
Capacity follows traffic.
Scale capacity around demand instead of provisioning infrastructure for peak traffic.
Incoming traffic
- Requests / s
- 120
- Active Pods
- 1
- GPU utilization
- 58%
- Queue
- 0
- Tokens / s
- 9.6K
Router · TR-IST-01
Baseline · 1 Pod
001
ACTIVE
002
STANDBY
003
STANDBY
004
STANDBY
005
STANDBY
006
STANDBY
007
STANDBY
008
STANDBY
009
STANDBY
010
STANDBY
011
STANDBY
012
STANDBY
013
STANDBY
014
STANDBY
015
STANDBY
016
STANDBY
Active
Warm
Standby
Illustrative simulation
End of Home 1 · Hero and system
Next — 05 / Models →