Every update. Every model.
Shipped in the open.

GPT-5.5 + Gemini 3.1 Pro now in the synthesis pool

We've added two flagship frontier models to Omniqra's reasoning pool. Every question now benefits from the latest releases — automatically routed based on prompt complexity, with no model switching required.

• 🚀 GPT-5.5 — OpenAI's most capable reasoning model
• 🧬 Gemini 3.1 Pro — Google's next-gen multimodal model
• ⚡ Average answer latency reduced by 18%
• 🎯 Synthesis accuracy improved on benchmark by +12.4%

Credits packs & pay-as-you-go billing

Launched a new credits system with 10+ pack sizes — from 25 credits for casual use, up to 50,000+ for enterprise teams. Every question uses just 1 credit, no matter how many models run behind the scenes.

• 💳 12 pack tiers with bulk discounts up to 40%
• 🎁 Free 3 credits/day for all signed-in users
• 🏢 Enterprise plans with custom pricing

Faster synthesis — answers in under 4 seconds

We rebuilt the synthesizer pipeline with parallel candidate streaming. Most queries now resolve in 2.8–3.9 seconds, down from 6+ seconds.

• ⚡ 38% faster average response time
• 🔄 Streaming partial answers as they arrive
• 📊 New live status indicator in the chat UI

Mobile-optimized chat & redesigned sidebar

Brand new mobile experience with a single floating OmniLogo, collapsible history, and improved keyboard handling for iOS & Android.

• 📱 Touch-optimized chat composer
• 🧭 Sleeker sidebar with thread search
• 🌙 Improved dark mode contrast across the app

Row-Level Security audit & hardened auth flow

Completed a full RLS audit across all database tables. Sessions are now scoped per user with refresh-token rotation and stricter CORS.

• 🔒 RLS enabled on 100% of user-data tables
• 🔁 Auth token rotation every 60 minutes
• 🛡 Added Cloudflare bot-shield on public endpoints

Markdown rendering & code-block fixes

Resolved multiple long-standing rendering issues in the answer area.

• ✅ Fixed nested list indentation
• ✅ Fixed code-block overflow on mobile
• ✅ Added language detection for 40+ programming languages
• ✅ One-click copy on every code block

Claude 3.5 Sonnet integrated into the synthesis pool

Anthropic's Claude 3.5 Sonnet joined the pool, dramatically improving answers requiring long-context reasoning, nuance and writing quality.

Compare page launched — Omniqra vs ChatGPT, Claude & Gemini

A new public comparison hub showing how Omniqra's multi-model synthesis outperforms single-vendor AI tools on real benchmarks.

Omniqra 2.0 — Multi-model synthesis goes public 🎉

The biggest release yet. Omniqra 2.0 introduces our proprietary 3-call synthesis engine: 2 frontier models run in parallel, then a third model synthesizes the best of both into one perfect answer.

• 🧠 Two candidates + one synthesizer = one truth
• 🎨 Brand refresh — new logo, dark mystical theme
• 📊 Public roadmap launched

Thread history with semantic search

You can now search across all your past chats by meaning, not just keywords.

Google sign-in & magic-link auth

One-tap sign-in with Google, plus password-less magic links via email.

🚀 Omniqra 1.0 — Public launch

The day it all started. Omniqra goes live to the public with its mission: "Don't ask one AI. Ask them all — and get the best answer."