New richer, animated charts on the Usage page: the requests-over-time bar chart has smooth grow animations, and the spend chart shows a green gradient area with a precise crosshair tooltip. A new model breakdown ring chart shows your top 5 models by request count and spend — derived from the logs already on screen, no extra loading.
The overview page sparkline is now interactive — hover any point to see the exact cost of that recent request.
Team API keys now have an hourly safety ceiling — a leaked key can no longer drain the entire team credit pool at unlimited request rate.
The playground streaming hook now stops cleanly when you navigate away, so a background chat can’t keep consuming credits.
Regenerating the last message in the playground now works correctly instead of resending the same turn.
Fixed the CSV usage export to include only the rows you filtered on (status, model) instead of always exporting everything.
Added structured breadcrumb metadata to Pricing, Privacy, Terms, Changelog, Support, and Status pages for better search-engine rich results.
Removed 30+ unused legacy components that were inflating the build — the production bundle is leaner.
Fixed an edge case where a slow payment webhook could lose credits if the checkout session expired before the credit grant arrived.
1.3.9 · Accessibility & reliability improvements
Numbers and prices throughout the dashboard and site now align correctly in their columns — token counts, spend figures, uptime percentages, and timestamps no longer shift width as they update.
Keyboard navigation improved across the whole site: pricing calculator sliders, For Agents language switcher, support quick-links, and dashboard sidebar links now all show a visible focus ring.
Screen readers now hear live updates from the pricing calculator, model search, status page refresh, settings save confirmation, and error banners.
Section anchors on Privacy and Terms pages now scroll to the correct position on deep-link, clearing the sticky header.
Model names, code tokens, and the NexGate wordmark on the homepage, models page, and For Agents page are now protected from auto-translation.
Header scroll transition now animates only the relevant CSS properties, reducing paint work on scroll.
1.3.7 · UI polish & bug fixes
Restyled the homepage model marquee to match the For Agents page — cleaner rounded pills with a soft fade-out on both edges.
Fixed a console error on buttons that render as links across the Header, Hero, and Call to Action sections.
1.3.6 · Marketing pages refresh
Refreshed the Models, Pricing, and For Agents pages with a clearer, more readable design that matches the new homepage.
Fixed pricing on phones — the Pricing, Models, and For Agents pages now show full pricing details on small screens instead of hiding columns.
Improved readability and contrast across the Support, Privacy, Terms, and Status pages, and kept the Status page navigation usable on mobile.
1.3.5 · Homepage & navigation redesign
Redesigned the homepage and navigation end to end — clearer, brighter, and more readable, with a right-sized headline and an animated feature grid that shows the product in action.
Rebuilt the mobile menu as a smooth slide-down panel with bigger tap targets, and the footer now animates cleanly instead of stuttering.
Fixed the FAQ section so it displays correctly on phones.
Fixed footer and menu links that pointed to the wrong place — API Reference, Status, and Contact now go where they should.
1.3.4 · Homepage redesign
Rebuilt the homepage for a cleaner flow and a faster load, with a redesigned hero that now looks right on phones as well as desktop.
Added a "How it works" section that walks you through getting started in three steps — grab a key, swap one line, ship requests.
Fixed the "Features" link in the footer, which previously reloaded the page instead of jumping to the features section.
1.3.3 · Design refresh
Refreshed the look across the site and dashboard — cleaner headings, a redesigned homepage, and a more polished sidebar on both desktop and mobile.
You can now create a team in one step right from the Team page — just type a name. No more hunting through menus.
The cost estimator on the pricing page now uses a simple model dropdown.
Fixed a hover glitch on the playground toolbar and tidied up tooltips throughout.
1.3.2 · Security & billing hardening
API key spending limits are now enforced — a key with a lifetime spend limit stops working once it reaches that cap, so keys you share stay within the limit you set.
Hardened sign-up abuse protection and made several billing and payment safeguards more reliable. No action needed on your part.
1.3.1 · Reliability & polish
Every available model responds reliably again, including DeepSeek, Grok, Kimi, and Llama.
Smoother playground: replies stream without lag in longer chats, and you can now select and copy text directly from a model’s response.
Each credit pack on the Top-up page shows how many are still available, so you can see when one is running low or sold out.
API key cards now show the real request count and spend per key.
The status page now flags when a refresh fails instead of quietly showing old data, and its uptime history scrolls cleanly on phones.
Removed the bright focus outline that appeared around inputs and buttons across the app and docs.
Retired the GPT-5.4 Pro model; all other GPT-5.4 models remain available.
1.3.0 · Platform overhaul
Manage multiple API keys with names, expiry dates, lifetime spend limits, and per-hour spend and request limits — and edit, rotate, or revoke them anytime.
New keyless playground: run any model from the dashboard with no API key. Usage is charged to your account balance.
Team roles (owner, admin, member): invite members by email, share a team credit pool, and let owners view team billing.
Redesigned the entire dashboard and the For Agents page with a cleaner, mobile-friendly layout.
Every model now works reliably for both streaming and non-streaming requests.
1.2.0 · Streamlined models and updated pricing
Cleaned up the model list by removing redundant variants — all capabilities remain available through the equivalent models.
Updated pricing on several models to current rates; Grok 4.3 and Grok 4 20 Reasoning are now cheaper to use.
1.1.0 · Multiple API keys and new models
Create multiple named API keys with custom expiry dates and spending limits.
Added the DeepSeek V4 Pro and Grok 4.3 models.
Rebuilt the developer playground with live token and cost estimates, model search, and reusable prompt templates.
1.0.0 · Documentation redesign
Added icons to every documentation page and section in the sidebar.
Added NEW badges to recently added pages.
Added quick-access navigation tabs to the documentation header (Docs, API Reference, Changelog, Models, Pricing).
Added a sidebar footer with direct links to llms-full.txt, the website, support, and the dashboard.
0.9.9 · AI Coding Assistants portal
Rebuilt /for-agents as a dedicated developer portal for AI coding assistants.
Added an interactive system prompt builder — select features and copy a ready-to-paste prompt.
Added tabbed code examples for Python, TypeScript, and cURL.
Added a live model reference table showing all available models and their tiers.
0.9.8 · Legal pages in docs
Privacy Policy and Terms of Service are now available inside /docs with full-text search and table of contents.
0.9.7 · Documentation light/dark mode
Documentation pages now have an independent light/dark mode toggle separate from the rest of the site.
0.9.5 · Clean dashboard URLs
Dashboard URLs are now clean on dashboard.nexgate.app — the address bar shows /usage, /topup, /settings instead of /dashboard/usage, /dashboard/topup, etc.
0.9.4 · API subdomain
The NexGate API is now available at api.nexgate.app/v1.
0.9.3 · Public status page
Added a public status page at status.nexgate.app.
90-day uptime bars and per-service health indicators.
Live incidents timeline with status and resolution history.
0.9.2 · Dashboard fixes
Fixed API key generation failing unexpectedly.
Fixed a crash when opening the quick-links menu in the dashboard topbar.
Improved legibility of sidebar footer links.
0.9.1 · Dashboard layout improvements
Settings page now uses a two-column layout on wider screens.
Overview and Usage tables hide less-important columns on smaller screens.
0.9.0 · Dashboard redesign
Rebuilt the dashboard with a new layout, improved design, and better accessibility.
Fixed focus handling in dialogs and drawers on mobile.
0.8.0 · Organizations
Added Organizations: create a team, invite members, and share a credit balance.
Team API keys are prefixed with ngo- and draw credits from the organization balance.
Organization owners can manage members, roles, invitations, and settings.
Improved bot protection on sign-up.
0.7.1 · In-app documentation
Documentation is now available at /docs with full-text search, sidebar navigation, and an API reference.
0.7.0 · Launch
NexGate is live.
Purchase prepaid credits and get an API key from the dashboard.
Use the API at /api/v1/chat/completions with any OpenAI-compatible client.
0.6.5 · Cost estimator and mobile polish
Improved cost estimator reliability on the pricing page.
Polished mobile layouts on /models and /for-agents.
0.6.4 · Rate limiting and request validation
Added rate limiting on public endpoints.
Improved API request validation before credit deduction.
0.6.1 · Streaming protection and health endpoint
Added a credit ceiling check before streaming starts to prevent overdraft on long responses.
Added a public /api/health endpoint.
0.6.0 · Models, pricing, and agent pages
Added /models — a public catalog of all available models.
Added /for-agents — integration guide for AI coding assistants.