■ 6 models · USD per 1M tokens

Models & pricing.

Up to 75% below official list prices, with zero data retention on every model. No minimums, no subscriptions.

  • DeepSeek V4.1 Flash

    DeepSeek · 1M context

    Save up to 75%
    ZDR
    Input
    $0.09$0.30 −70%
    Output
    $0.30$1.20 −75%
    Cache read
    $0.006

    deepseek/deepseek-v4.1-flash

  • GLM 5.3 Flash

    Z.ai · 1M context

    Save up to 40%
    ZDR
    Input
    $0.10$0.15 −33%
    Output
    $0.30$0.50 −40%
    Cache read
    $0.025$0.03 −17%

    zai/glm-5.3-flash

  • GLM 5.3 Flash Uncensored

    Z.ai · 1M context

    ZDRUncensored
    Input
    $0.18
    Output
    $0.72
    Cache read
    $0.063

    zai/glm-5.3-flash-uncensored

  • MiMo V2.6 Pro

    Xiaomi · 1M context

    ZDR
    Input
    $0.43
    Output
    $0.87
    Cache read
    $0.0036

    xiaomi/mimo-v2.6-pro

  • GLM 5.3

    Z.ai · 1M context

    Save up to 75%
    ZDR
    Input
    $0.60$1.40 −57%
    Output
    $1.10$4.40 −75%
    Cache read
    $0.20$0.26 −23%

    zai/glm-5.3

  • GLM 5.3 Uncensored

    Z.ai · 1M context

    ZDRUncensored
    Input
    $1.125
    Output
    $2.025
    Cache read
    $0.27

    zai/glm-5.3-uncensored

Every model is served with zero data retention (ZDR). Struck-through prices are each developer's own API list price as of October 2026; DeepSeek is compared at its peak-hour rate.

■ Calculator

Estimate your bill.

Estimated monthly cost

$0.75

Input
$0.45
Cache read
$0.00
Output
$0.30

■ You save $1.95 vs $2.70 at official prices.

■ FAQ

Billing questions.

How am I billed?+

You're charged per token, separately for input (prompt) and output (completion) tokens, at the per-million rates listed above. Input tokens served from the prompt cache are billed at the lower cache-read rate. Usage is metered in real time and billed from your prepaid balance.

Do you store my prompts?+

No. Every model on Conduit runs with zero data retention (ZDR): prompts and completions are never stored or used for training.

How are the discounts calculated?+

We compare our rates with each model developer's own API list price as of October 2026. DeepSeek V4.1 Flash is compared at DeepSeek's peak-hour rate.

Are there minimums or subscriptions?+

No. There are no monthly fees, seat licenses or commitments. Top up your balance and pay only for what you use.

Do reasoning tokens count as output?+

Yes. For reasoning models, thinking tokens are billed at the output rate, and they're reported separately in the usage object of every response.

Can I get volume discounts?+

Teams processing more than a few billion tokens a month can contact Enverth Research for committed-use pricing and dedicated capacity.