{"id":315,"date":"2026-07-25T22:05:08","date_gmt":"2026-07-25T22:05:08","guid":{"rendered":"https:\/\/efiwe.com\/blog\/?p=315"},"modified":"2026-07-25T22:23:43","modified_gmt":"2026-07-25T22:23:43","slug":"efiwe-learners-can-now-choose-their-own-ai-tutor","status":"publish","type":"post","link":"https:\/\/efiwe.com\/blog\/efiwe-learners-can-now-choose-their-own-ai-tutor\/","title":{"rendered":"Efiwe Learners Can Now Choose Their Own AI Tutor"},"content":{"rendered":"\n

The mobile-first coding education platform hands learners direct control over which AI model reviews their code, turning a choice that once required switching off the internet into a single tap; on the web platform and in the JavaScript SDK alike.<\/em><\/p>\n\n\n

\n
\"\"<\/figure>\n<\/div>\n\n\n

Efiwe, the mobile-first coding education platform, has introduced an AI model selector that lets every learner decide which of its two AI tutors reviews their work. A new dropdown sits directly beneath the challenge controls, offering a straightforward choice: Efiwe 2.0<\/strong>, the cloud model, or Efiwe 1.5<\/strong>, the on-device model that runs entirely offline. The setting takes effect immediately and applies to every piece of feedback the platform generates, from code validation to hints to issue diagnoses.<\/p>\n\n\n\n

Until now, that choice existed only by accident. Efiwe’s dual-model architecture has always fallen back to the on-device model when the cloud was unreachable, which meant a learner who preferred the offline tutor had exactly one way to get it: switch off their WiFi or mobile data. It worked, but it was a workaround rather than a feature, and it forced learners to disconnect their entire device to change a single setting inside one app.<\/p>\n\n\n\n

The new selector makes the preference explicit. A learner who chooses Efiwe 1.5 gets the on-device model every time, regardless of how strong their connection is. A learner who leaves the default alone gets exactly the behaviour they have always had.<\/p>\n\n\n\n

What each option actually does<\/strong><\/h2>\n\n\n\n

Choosing Efiwe 2.0<\/strong> changes nothing. Free learners continue to receive cloud-powered feedback once every five minutes, with the on-device model handling everything in between. Efiwe Pro subscribers and members of Group Pro and Group Pro Plus plans continue to reach the cloud model on every submission, with the on-device model stepping in only when the cloud is busy or unavailable. This remains the default for every learner, and no one needs to touch the dropdown to keep it.<\/p>\n\n\n\n

Choosing Efiwe 1.5<\/strong> routes every request to the device. Nothing is sent to the cloud, no connection is consulted, and feedback arrives at the speed of local computation. For free learners this carries a quiet benefit: the five-minute cooldown exists to keep the cloud service sustainable, so it simply does not apply. A learner on Efiwe 1.5 gets instant, unlimited feedback on every challenge, with no waiting between attempts and no prompts to upgrade.<\/p>\n\n\n\n

That trade is a real one, and it is worth being honest about both directions. Efiwe 2.0 reasons about intent, diagnoses the gap between what a learner wrote and what the challenge wanted, and explains mistakes in beginner-friendly prose. Efiwe 1.5, sharpened further by the 1.5.1 release earlier this month, catches typos down to the exact letters at fault, flags wrong units and operators, checks indentation and missing colons, and judges close-enough answers fairly. It is faster and it is private, but it explains less.<\/p>\n\n\n\n

Why learners were asking for this<\/strong><\/h2>\n\n\n\n

The learners who pushed hardest for the option were not the ones with the worst connections. They were the ones who had learned to prefer the offline model on its own merits.<\/p>\n\n\n\n

“We built the fallback because connections drop, and we assumed that was the only reason anyone would ever land on the on-device model,”<\/strong> said Chidi Nwaogu, Co-founder and CTO of Efiwe. “Then learners started telling us they were switching off their data on purpose. Some wanted the speed. Some were watching every megabyte on a prepaid bundle. Some just did not want their code leaving their phone. Those are all good reasons, and none of them should require turning off your internet to act on.”<\/strong><\/p>\n\n\n\n

The data-cost point matters more than it might appear from a well-connected city. For learners on metered prepaid bundles, an AI feature that quietly consumes data on every submission is a feature they may ration or avoid. Making the on-device model a deliberate, sticky choice means those learners can use Efiwe as heavily as they like without watching a balance drain.<\/p>\n\n\n\n

The choice follows the learner<\/strong><\/h2>\n\n\n\n

The preference is saved to the learner’s Efiwe account rather than to a single course or a single phone. Set it once in the HTML course and it holds in CSS, JavaScript, Tailwind, Python, and SQL. Sign in on a shared classroom device or a replacement handset and it follows, restored the moment the account loads. For the many Efiwe learners who share a phone with siblings or classmates, this means the setting travels with the person, not the hardware.<\/p>\n\n\n\n

Available in the SDK too<\/strong><\/h2>\n\n\n\n

The selector is not limited to efiwe.com<\/strong>. It ships in the Efiwe JavaScript SDK as well, so learners using Efiwe embedded inside a partner school’s platform, an NGO’s app, or any of the products built on the Efiwe Open API see the same dropdown and get the same control. In the SDK the choice is remembered in the learner’s browser, and it governs feedback, hints, and diagnoses exactly as it does on the main platform.<\/p>\n\n\n\n

This is consistent with how Efiwe has approached the API from the start: partners who embed Efiwe should be handing their learners the full platform, not a reduced version of it. When a capability reaches learners on efiwe.com, it reaches learners everywhere Efiwe runs.<\/p>\n\n\n\n

Why the switch?<\/strong><\/h2>\n\n\n\n

Efiwe’s dual-model architecture has always been built on the belief that no learner should have to choose between smart feedback and offline capability. The model selector extends that belief one step further. Having built both models to be genuinely great, the platform is now confident enough to let learners decide for themselves which one fits their connection, their data budget, their privacy preferences, and the way they like to learn.<\/p>\n\n\n\n

The AI model selector is available now on efiwe.com<\/strong> and in the Efiwe JavaScript SDK<\/strong>. It requires no update or setup, and Efiwe 2.0 remains the default for every learner who does not change it.<\/p>\n","protected":false},"excerpt":{"rendered":"

Choosing Efiwe 2.0 changes nothing. Free learners continue to receive cloud-powered feedback once every five minutes, with the on-device model handling everything in between. Efiwe Pro subscribers and members of Group Pro and Group Pro Plus plans continue to reach the cloud model on every submission, with the on-device model stepping in only when the cloud is busy or unavailable. This remains the default for every learner, and no one needs to touch the dropdown to keep it. Choosing Efiwe 1.5 routes every request to the device. Nothing is sent to the cloud, no connection is consulted, and feedback arrives at the speed of local computation. For free learners this carries a quiet benefit: the five-minute cooldown exists to keep the cloud service sustainable, so it simply does not apply. A learner on Efiwe 1.5 gets instant, unlimited feedback on every challenge, with no waiting between attempts and no prompts to upgrade.<\/p>\n","protected":false},"author":1,"featured_media":318,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[17,79,292,14,291,4,293,229,228,120,290],"class_list":["post-315","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized","tag-1-0","tag-1-5","tag-1-5-1","tag-2-0","tag-default","tag-efiwe","tag-efiwe-1-0","tag-efiwe-1-5","tag-efiwe-1-5-1","tag-efiwe-2-0","tag-selector"],"_links":{"self":[{"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/posts\/315","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/comments?post=315"}],"version-history":[{"count":4,"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/posts\/315\/revisions"}],"predecessor-version":[{"id":320,"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/posts\/315\/revisions\/320"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/media\/318"}],"wp:attachment":[{"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/media?parent=315"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/categories?post=315"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/efiwe.com\/blog\/wp-json\/wp\/v2\/tags?post=315"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}