<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>model-gateway on tomrochette.com</title>
    <link>https://tomrochette.com/tags/model-gateway/</link>
    <description>Recent content in model-gateway on tomrochette.com</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en</language>
    <managingEditor>tom@tomrochette.com (Tom Rochette)</managingEditor>
    <webMaster>tom@tomrochette.com (Tom Rochette)</webMaster>
    <copyright>© 2026 Tom Rochette</copyright>
    <lastBuildDate>Sat, 26 Sep 2026 03:01:49 -0400</lastBuildDate><atom:link href="https://tomrochette.com/tags/model-gateway/index.xml" rel="self" type="application/rss+xml" />
    
    <item>
      <title>NanoGPT</title>
      <link>https://tomrochette.com/agents/model-access/nanogpt/</link>
      <pubDate>Sat, 26 Sep 2026 00:00:00 +0000</pubDate>
      <author>tom@tomrochette.com (Tom Rochette)</author>
      <guid>https://tomrochette.com/agents/model-access/nanogpt/</guid>
      <category>research-note</category><category>agent-curated</category><category>fully-ai-generated</category><category>llm=glm-5.3-flash</category><category>model-access</category><category>model-gateway</category><category>open-weights</category><category>privacy</category>
      <description>&lt;p&gt;NanoGPT (nano-gpt.com) is an independent model-access gateway combining pay-per-token API access at list prices with a flat subscription that bundles open-weight model usage at no per-token cost.&#xA;&lt;strong&gt;At $12 per month it is the cheapest flat-rate open-weight tier I can verify anywhere, but the deal quietly shrank over the past year.&lt;/strong&gt;&lt;/p&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;What it is&#xA;    &lt;div id=&#34;what-it-is&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#what-it-is&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;p&gt;The service runs a hosted web chat and an OpenAI-compatible API at provider list prices with no markup, taking deposits from $0.10 in crypto or $1 by card.&#xA;The Pro subscription includes open-weight text and image models, currently framed as 60 million input tokens per week plus a 5% discount on eligible paid text models.&#xA;Coverage spans text, image, video, 3D, audio, and embedding models under one balance.&#xA;An independent operator runs it, and the pages I fetched name no corporate parent.&lt;/p&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Status&#xA;    &lt;div id=&#34;status&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#status&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;p&gt;Active: the blog published within the week of verification (September 22, 2026), ships monthly crypto-payment statistics, and added zero-data-retention routing on September 3, 2026.&#xA;The Wayback Machine holds 57 captures from September 2024 through September 2026, so the service has operated continuously for at least two years.&#xA;Its own Hacker News footprint is nearly nil: two stories in late 2024 with zero comments, while the NanoGPT name on HN belongs to Karpathy&amp;rsquo;s training repo (a 1,532-point story).&lt;/p&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Strengths&#xA;    &lt;div id=&#34;strengths&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#strengths&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;ul&gt;&#xA;&lt;li&gt;List-price API with the exact cost printed on every request makes it a rare auditable gateway.&lt;/li&gt;&#xA;&lt;li&gt;The subscription undercuts every first-party lab plan while covering text and image generation.&lt;/li&gt;&#xA;&lt;li&gt;The privacy posture is concrete: crypto payments (Monero led deposits at 38.55% in August 2026), no-prompt-logging claims, and ZDR routing.&lt;/li&gt;&#xA;&lt;/ul&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Cautions&#xA;    &lt;div id=&#34;cautions&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#cautions&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;ul&gt;&#xA;&lt;li&gt;The deal narrowed: September 2025 sold unlimited personal open-weight use at $8, September 2026 sells 60M input tokens per week at $12, a 50% rise with a new cap.&lt;/li&gt;&#xA;&lt;li&gt;Bring-your-own-key requests are billed at 5% of normal model cost, and pinning a specific provider adds another 5%.&lt;/li&gt;&#xA;&lt;li&gt;Governance is opaque: no funding, ownership, or company facts appear on fetched pages, so nobody is accountable if terms move again.&lt;/li&gt;&#xA;&lt;li&gt;Scrutiny lives in Discord and niche communities rather than HN, so fewer independent eyes check the claims.&lt;/li&gt;&#xA;&lt;/ul&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Pricing&#xA;    &lt;div id=&#34;pricing&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#pricing&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;p&gt;Pro is $12/month as of 2026-09-26, including 60M input tokens per week, the 5% paid-text-model discount, and web plus API access.&#xA;Pay-as-you-go needs no subscription: a free tier with one web-only model, list-price API billing, and $0.10 crypto or $1 card deposit minimums.&#xA;In September 2025, Pro was $8/month with unlimited personal open-weight usage capped at 60,000 generations per month and 2,000 per day.&lt;/p&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Price history&#xA;    &lt;div id=&#34;price-history&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#price-history&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;table&gt;&#xA;&#x9;&lt;thead&gt;&#xA;&#x9;&#x9;&#x9;&lt;tr&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;th&gt;Date&lt;/th&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;th&gt;Plan&lt;/th&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;th&gt;Change&lt;/th&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;th&gt;Source&lt;/th&gt;&#xA;&#x9;&#x9;&#x9;&lt;/tr&gt;&#xA;&#x9;&lt;/thead&gt;&#xA;&#x9;&lt;tbody&gt;&#xA;&#x9;&#x9;&#x9;&lt;tr&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;2025-09-01&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;Pro&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;$8/month, unlimited personal open-weight usage (60,000 generations/month, 2,000/day)&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;&lt;a href=&#34;http://web.archive.org/web/20250901063827/https://nano-gpt.com/subscription&#34;  target=&#34;_blank&#34; rel=&#34;noreferrer&#34;&gt;&lt;img class=&#34;external-link-favicon&#34; src=&#34;https://www.google.com/s2/favicons?domain=web.archive.org&amp;sz=128&#34; alt=&#34;&#34; width=&#34;16&#34; height=&#34;16&#34; loading=&#34;lazy&#34;&gt;http://web.archive.org/web/20250901063827/https://nano-gpt.com/subscription&lt;/a&gt;&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&lt;/tr&gt;&#xA;&#x9;&#x9;&#x9;&lt;tr&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;2026-09-26&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;Pro&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;$12/month, 60M included input tokens per week, 5% paid-text-model discount&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&#x9;&#x9;&lt;td&gt;&lt;a href=&#34;https://nano-gpt.com/pricing&#34;  target=&#34;_blank&#34; rel=&#34;noreferrer&#34;&gt;&lt;img class=&#34;external-link-favicon&#34; src=&#34;https://www.google.com/s2/favicons?domain=nano-gpt.com&amp;sz=128&#34; alt=&#34;&#34; width=&#34;16&#34; height=&#34;16&#34; loading=&#34;lazy&#34;&gt;https://nano-gpt.com/pricing&lt;/a&gt;&lt;/td&gt;&#xA;&#x9;&#x9;&#x9;&lt;/tr&gt;&#xA;&#x9;&lt;/tbody&gt;&#xA;&lt;/table&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Compared to&#xA;    &lt;div id=&#34;compared-to&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#compared-to&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;p&gt;Kimi Code (../kimi-code/index.md) is the pick when you want one lab&amp;rsquo;s frontier open-weight models with quota predictability at a similar monthly price.&#xA;MiniMax Coding Plan (../minimax-coding-plan/index.md) covers multimodal token volume cheaply but locks you out in 5-hour windows, which NanoGPT never does.&#xA;The free PAYG path also makes NanoGPT the lowest-commitment way to test a model before any subscription.&lt;/p&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Bottom line&#xA;    &lt;div id=&#34;bottom-line&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#bottom-line&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;p&gt;Recommended for privacy-minded tinkerers and open-weight power users who want flat-cost breadth without quota-window lockouts.&#xA;Not for engineers whose work needs closed frontier models at subscription quotas, since the $12 tier&amp;rsquo;s value lives almost entirely in open weights.&#xA;I will claim something arguable: at $12 with open weights included, this beats any $15-$20 first-party plan on value, because open-weight quality has closed enough of the gap that the labs&amp;rsquo; premium is now a convenience fee.&lt;/p&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;Changes&#xA;    &lt;div id=&#34;changes&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#changes&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;ul&gt;&#xA;&lt;li&gt;2026-09-26 - Created.&lt;/li&gt;&#xA;&lt;/ul&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;See also&#xA;    &lt;div id=&#34;see-also&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#see-also&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;ul&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://tomrochette.com/agents/model-access/kimi-code/&#34; &gt;Kimi Code&lt;/a&gt; - the first-party open-weight subscription to compare quota mechanics against.&lt;/li&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://tomrochette.com/agents/model-access/minimax-coding-plan/&#34; &gt;MiniMax Coding Plan&lt;/a&gt; - the token-metered alternative with modality breadth.&lt;/li&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://tomrochette.com/agents/model-provider-feature-matrix/&#34; &gt;Model provider feature matrix&lt;/a&gt; - where gateways sit against first-party plans.&lt;/li&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://tomrochette.com/agents/model-selection-for-coding-tasks/&#34; &gt;Model selection for coding tasks&lt;/a&gt; - choose models on evidence before choosing where to buy them.&lt;/li&gt;&#xA;&lt;/ul&gt;&#xA;&#xA;&lt;h2 class=&#34;relative group&#34;&gt;References&#xA;    &lt;div id=&#34;references&#34; class=&#34;anchor&#34;&gt;&lt;/div&gt;&#xA;    &#xA;    &lt;span&#xA;        class=&#34;absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none&#34;&gt;&#xA;        &lt;a class=&#34;text-primary-300 dark:text-neutral-700 !no-underline&#34; href=&#34;#references&#34; aria-label=&#34;Anchor&#34;&gt;#&lt;/a&gt;&#xA;    &lt;/span&gt;&#xA;    &#xA;&lt;/h2&gt;&#xA;&lt;ul&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://nano-gpt.com/pricing&#34;  target=&#34;_blank&#34; rel=&#34;noreferrer&#34;&gt;&lt;img class=&#34;external-link-favicon&#34; src=&#34;https://www.google.com/s2/favicons?domain=nano-gpt.com&amp;sz=128&#34; alt=&#34;&#34; width=&#34;16&#34; height=&#34;16&#34; loading=&#34;lazy&#34;&gt;https://nano-gpt.com/pricing&lt;/a&gt; - current Pro pricing ($12/month, 60M input tokens/week), PAYG minimums, 5% BYOK and pinning fees, as of 2026-09-26.&lt;/li&gt;&#xA;&lt;li&gt;&lt;a href=&#34;http://web.archive.org/web/20250901063827/https://nano-gpt.com/subscription&#34;  target=&#34;_blank&#34; rel=&#34;noreferrer&#34;&gt;&lt;img class=&#34;external-link-favicon&#34; src=&#34;https://www.google.com/s2/favicons?domain=web.archive.org&amp;sz=128&#34; alt=&#34;&#34; width=&#34;16&#34; height=&#34;16&#34; loading=&#34;lazy&#34;&gt;http://web.archive.org/web/20250901063827/https://nano-gpt.com/subscription&lt;/a&gt; - September 2025 capture showing $8/month with unlimited personal open-weight usage.&lt;/li&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://nano-gpt.com/blog&#34;  target=&#34;_blank&#34; rel=&#34;noreferrer&#34;&gt;&lt;img class=&#34;external-link-favicon&#34; src=&#34;https://www.google.com/s2/favicons?domain=nano-gpt.com&amp;sz=128&#34; alt=&#34;&#34; width=&#34;16&#34; height=&#34;16&#34; loading=&#34;lazy&#34;&gt;https://nano-gpt.com/blog&lt;/a&gt; - activity evidence: September 2026 posts, ZDR routing, monthly crypto-payment statistics.&lt;/li&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://hn.algolia.com/api/v1/search?query=nano-gpt.com&amp;amp;restrictSearchableAttributes=url&amp;amp;tags=story&amp;amp;hitsPerPage=10&#34;  target=&#34;_blank&#34; rel=&#34;noreferrer&#34;&gt;&lt;img class=&#34;external-link-favicon&#34; src=&#34;https://www.google.com/s2/favicons?domain=hn.algolia.com&amp;sz=128&#34; alt=&#34;&#34; width=&#34;16&#34; height=&#34;16&#34; loading=&#34;lazy&#34;&gt;https://hn.algolia.com/api/v1/search?query=nano-gpt.com&amp;restrictSearchableAttributes=url&amp;tags=story&amp;hitsPerPage=10&lt;/a&gt; - the service&amp;rsquo;s thin HN footprint (two 0-comment stories, 2024).&lt;/li&gt;&#xA;&lt;li&gt;&lt;a href=&#34;https://hn.algolia.com/api/v1/search?query=NanoGPT&amp;amp;tags=story&amp;amp;hitsPerPage=10&#34;  target=&#34;_blank&#34; rel=&#34;noreferrer&#34;&gt;&lt;img class=&#34;external-link-favicon&#34; src=&#34;https://www.google.com/s2/favicons?domain=hn.algolia.com&amp;sz=128&#34; alt=&#34;&#34; width=&#34;16&#34; height=&#34;16&#34; loading=&#34;lazy&#34;&gt;https://hn.algolia.com/api/v1/search?query=NanoGPT&amp;tags=story&amp;hitsPerPage=10&lt;/a&gt; - the name collision with Karpathy&amp;rsquo;s repo dominating HN results.&lt;/li&gt;&#xA;&lt;/ul&gt;&#xA;</description>
      
    </item>
    
  </channel>
</rss>
