V26 Inference
V26 Inference — Affordable AI inference for developers

Affordable AI inference
for developers.

V26 provides affordable AI inference for developers, starting with Qwen3.8-27B — delivered through a simple, OpenAI-compatible API.

Join the Beta — Get Free Credits → Explore Qwen3.8-27B
Qwen3.8-27B API — Coming Soon

Join the beta to receive free credits when the API launches.

Developer focused OpenAI-compatible Free beta credits Affordable inference
RequestClient
RouterInference
Qwen3.8-27BModel
ResponseAPI

Planned request path

01 / Platform

Built for developers who just want to use AI.

Simple API

The planned API will use familiar OpenAI-compatible request patterns so existing apps adopt V26 with minimal changes.

InterfaceOpenAI-compatible

Affordable by design

V26 is focused on efficient inference and transparent limits. Pricing will be announced before public beta.

PricingAnnounced pre-beta

Free to try

Early developers will receive controlled free beta credits so we can learn what workloads matter most.

AccessControlled beta credits

Developer first

Designed around developers: clear docs, usage visibility, and simple onboarding. Additional open models may be supported in the future.

FocusDeveloper-first
02 / Model
Launching first · Coming soon

Qwen3.8-27B

The first model planned for V26. API access is coming soon — join the beta for free credits. V26 will offer an OpenAI-compatible interface and may support additional open models in the future.

Model
Qwen3.8-27B
Parameters
27B
Interface
OpenAI-compatible API
Status
Coming soon
Access
Beta / Free credits
Pricing
TBA — announced before beta
Join the Beta — Get Free Credits →

One model to start, chosen as a practical open-weight baseline for coding, professional work, research, and agents.

More models may follow as demand grows. Each addition will be documented with the same specification format.

Full benchmarks, context details, and pricing will be published before public beta.

03 / Architecture

Request in. Intelligence out.

ClientYour application
APIOpenAI-compatible · Planned
Inference RouterPlanned
Model SelectionPlanned
Qwen3.8-27BFirst model
Future modelsAs demand grows
GPU ClusterPlanned capacity
ResponseStandard format

V26 is designed around a single, legible request path: requests enter through an OpenAI-compatible API, route to the selected model, execute on GPU capacity, and return as a standard response.

Routing exists so future models can be added behind the same interface — existing integrations keep working while the fleet grows.

Planned architecture. Final topology, regions, and redundancy will be published before public beta.

04 / Developers

Bring your existing OpenAI SDK.

V26 is being built with an OpenAI-compatible API, making it easy to switch existing applications to V26.

Planned API
from openai import OpenAI

client = OpenAI(
    api_key="v26_...",
    base_url="https://api.v26-inference.com/v1"
)

response = client.chat.completions.create(
    model="qwen3.8-27b",
    messages=[
        {"role": "user",
         "content": "Hello from V26!"}
    ]
)

print(response.choices[0].message.content)
API endpoint shown for the planned V26 interface. The inference endpoint is not live yet.
05 / Access

Be one of the first V26 developers.

Tell us where you would use Qwen3.8-27B. We use early demand to shape capacity, limits and pricing. Pricing will be announced before public beta.

Qwen3.8-27B API access is coming soon. Join the beta and receive free credits when the API launches. Registration runs via Google Forms and takes ~1 minute — only email, use case, and expected monthly token usage are required. Name, company, current AI provider, country/region, and additional comments are optional. Consent is captured in the form. See Privacy and Terms.

Join the Beta — Get Free Credits →

Opens Google Forms in a new tab. No payment required. Free beta credits only — no specific allocation claimed. Prefer the full page? Open the registration page.

Know a developer who needs affordable AI inference? Share V26 with them — send them to v26-inference.com.

Questions, partnerships or early enterprise interest? hello@v26-inference.com

FAQ

Questions, answered honestly.

Is the API live today?

Not yet. V26 is validating demand before committing to production GPU capacity. Join the beta and we will notify you.

Why Qwen3.8-27B?

A practical open-weight starting point for coding, professional work, research and agents. More models may follow as demand grows.

Will there be a free tier?

Yes — a controlled free-credit tier. Exact limits and pricing will be announced before public beta.

Will V26 support OpenAI SDKs?

That is the planned interface: OpenAI-compatible request patterns. The beta endpoint is not live yet.

How do I contact you?

Email hello@v26-inference.com or use the contact page.