Skip to content

For authorized security work

Security answers, not refusals.

Open-weight models for red teams, pentesters, malware analysts and detection engineers. One API key, priced per token, with the cost of every request on the receipt.

Get an API keyRead the quickstart

For people 18 and over doing lawful work they are authorized to do. Read the acceptable use policy.

Models

Every model, one key.

One published price per model, in US dollars per 1M tokens. Fees included.

deepseek deepseek-v4.1-flash-uncensored
Coming soonkimikimi-k3.1-pro-uncensored
See all models
ModelContextInputOutput
DeepSeek V4.1 Flash (uncensored)deepseek
batchwell/deepseek-v4.1-flash-uncensored
1M$0.44$1.49
  • DeepSeek V4.1 Flash (uncensored)deepseek
    batchwell/deepseek-v4.1-flash-uncensored

    $0.44 in · $1.49 out per 1M tokens

    1M context

Most assistants decline anything that sounds like an attack. Ours are open-weight checkpoints with those refusals removed, so they stay useful on the engagements you are cleared for.

The work

Built for the engagement.

Useful when the topic sounds offensive and the work is authorized.

Red team and pentest

Plan the engagement, draft payload notes, write the report. No refusal because the topic sounds offensive.

Platform

One endpoint. No surprises.

The API you already use, with the routing and receipts you need.

OpenAI-compatible

Point the OpenAI SDK at our base URL. Chat completions, streaming and tool calls work as you expect.

API

Drop-in for the OpenAI SDK.

Change the base URL and the key. Keep the code you have.

Read the quickstart
curl https://api.batchwell.ai/v1/chat/completions \
  -H "Authorization: Bearer $BATCHWELL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "batchwell/deepseek-v4.1-flash-uncensored",
    "messages": [
      {
        "role": "system",
        "content": "Authorized internal pentest, scope ACME-2026-14."
      },
      {
        "role": "user",
        "content": "Which log sources would catch this Kerberoasting run?"
      }
    ]
  }'

Privacy

We keep the receipt, not the conversation.

Your prompts and completions are not stored.

Your request metadata stays, so you can audit spend.

Read the privacy notice

Billing

Prepaid. Refundable for 7 days.

7 days

Refund window

Top up from $20 to $2,000. Credit never expires. Prices already include our fees.

How billing works

1

Models in the catalogue

$0.44

Lowest price per 1M input tokens

$20

Smallest top-up

FAQ

Questions, answered.

Who can use Batchwell?

People 18 and over doing lawful work they are authorized to do, under the acceptable use policy.

Do you store my prompts?

No. We keep request metadata (model, token counts, cost, timing, status) so you can audit spend.

What is not allowed?

Anything the acceptable use policy forbids. Read it before you create an account.

Which API do you support?

The OpenAI Chat Completions API, with streaming and tool calls. Point the OpenAI SDK at https://api.batchwell.ai/v1.

How does billing work?

Prepaid credit, $20 to $2,000 per top-up. Credit never expires, and unused credit is refundable within 7 days of purchase.

What if a model fails?

If an upstream fails before the first token, we try your fallback models. Once text starts, we never switch silently.

Talk to us

Ask the question you actually have.

Open-weight models. OpenAI-compatible. Priced per token.

Get an API keyRead the quickstart