Anyone who has taken a taxi in Addis Ababa knows the conversation. You name your destination, the driver names a price, and the next two minutes are a negotiation. If it is raining, or it is rush hour at Megenagna, or you are arriving at Bole with luggage and a foreign accent, the number goes up. Ride-hailing apps were supposed to end this, but many of them simply moved the negotiation into an algorithm: surge pricing that rises with demand, and a final fare that can differ from the estimate.

When we built BinaRide, the ride service inside BinaSmart, we started from one rule. The fare is calculated when you book, it is locked, and it does not change. Not for traffic, not for rain, not because more people want a car at that moment.

What a locked fare looks like

Here is a real quote from our fare engine at the time of writing, for a trip from Bole International Airport to Piassa, about 9 kilometres:

• Economy (4 seats): 365 birr
• Comfort (4 seats): 525 birr
• Van (7 seats): 730 birr

You see these prices before you confirm. The driver sees the same pickup, drop-off, distance and earnings on the ride card before accepting. Nobody can change the number afterwards, so there is nothing to argue about at the end of the trip. You pay the driver in cash or with telebirr.

This matters most at the airport. A visitor landing at Terminal 2 at night has no idea what a fair price is, and that is exactly when a locked fare is most useful. It also means family abroad can help: someone in Dubai or Washington can book the ride for a parent arriving in Addis, and know exactly what it will cost.

Why drivers agreed to it

A fixed fare only works if drivers accept it, and drivers have good reasons to like surge pricing. What changed their minds was the other half of the deal.

During our launch, BinaSmart takes no commission. The whole fare goes to the driver, paid directly by the passenger. There is no sign-up fee, no weekly fee and no deposit, and drivers are free to keep using other apps. We have committed to telling drivers in writing before that ever changes.

Keeping 100% of a predictable fare is a different deal from keeping a share of an unpredictable one. And a driver who never has to argue with a passenger has a better day.

Sharing the car on the commute

The same fare engine powers BinaPool, our shared commute. Up to four riders share one Comfort car on a fixed corridor and pay per seat. The more people in the car, the less each person pays, and the more the driver earns.

On the Kazanchis to Piassa corridor, via Arat Kilo, the seat prices at the time of writing were:

• 1 rider: 295 birr
• 2 riders: 165 birr each
• 3 riders: 120 birr each
• 4 riders: 100 birr each

With four riders, each person pays about a third of the solo fare, and the driver takes 385 birr instead of 295. Everyone has a reason to want a full car. The corridors follow the city's real commuting patterns, such as Megenagna to Bole, CMC to Bole, and Piassa or Mexico to Kazanchis, running inbound in the morning and outbound in the afternoon. It is, in a sense, a formal version of the contract taxi arrangements that Addis has relied on for years.

Built for 3G, not for app stores

Our biggest design decision was not about pricing at all. BinaRide has no app to download. Riders book inside Telegram, through @bina_smart_bot, or in the browser at bina.et/ride. Drivers sign up and take trips inside Telegram too, through @binasmartdriverbot, and it works well on 3G.

Telegram is already one of the most used apps on Ethiopian phones. Asking someone to download a new app, on a slow connection and a phone with little storage, is where many good services lose their users. Meeting people where they already are removed that step entirely.

What we would tell other builders

• Make the price the promise. In a market used to negotiation, a number that cannot change is a feature in itself.
• Align the incentives. Pool pricing works because riders, drivers and the platform all want the same thing: a full car.
• Respect the network people actually have. If your service needs 4G and a fresh download, you have already excluded much of your market.
• Earn trust before you charge for it. Zero commission during launch let drivers try the model without risk.

BinaRide currently serves Addis Ababa, with Economy, Comfort and Van cars, and airport pickups at Bole Terminal 2. You can check a fixed price for your own trip at bina.et/ride, see the shared commute corridors at bina.et/pool, or book an airport transfer at bina.et/airport. Drivers can join at bina.et/drive-with-us.

Ibrahim Kedir Bedru is the founder of BinaSmart (bina.et), based at JJ Darule Building in Addis Ababa. Fares quoted are from BinaSmart's fare engine at the time of writing; the live price is always shown before you book.

Originally published on Medium.

Ask a general-purpose AI assistant how much it costs to renew an Ethiopian passport, or how to recover a lost Fayda number, and you will usually get a confident answer. Often it is out of date, and sometimes it is simply made up. The model learned about Ethiopia from whatever happened to be on the internet when it was trained, and Ethiopian government procedures change faster than those snapshots.

That was the problem we set out to solve at BinaSmart, an Amharic-first app for everyday life in Ethiopia. People come to us to book a ride in Addis Ababa or find a hotel, but they also ask practical questions: what documents a business licence needs, what the new salary tax bands are, which office handles an urgent passport. We wanted an assistant that could answer those questions in Amharic and be right.

Not a chatbot, a librarian

Bini, our assistant, doesn't answer from memory. On every question it searches an indexed library of Ethiopian documents (laws, proclamations, government service pages and official fee tables), reads the passage that matches, and answers from that passage. Every answer has a source you can open and check.

This design choice matters more than the choice of model. When the salary tax bands changed under Proclamation 1395/2025, we didn't need to retrain anything. We added the new proclamation to the library, and Bini started answering from it the same day.

On our public test of 138 real questions about government services, Bini currently answers 96% correctly. The questions it misses are useful too: each one points to a document we haven't indexed yet.

The Amharic token tax, measured

AI is billed in tokens, and Amharic is expensive in tokens. Our own news desk first measured Amharic at 6.1 times the tokens of the same text in English, using OpenAI's o200k_base tokenizer.

BinaSmart runs on Gemini, so we measured again with the tokenizer that actually governs our bill. The result was 1.9 times, not 6.1.

Both numbers are correct, and the gap is the lesson: the penalty belongs to the tokenizer, not to the Ge'ez script. A tokenizer that has learned Amharic splits it into sensible pieces, and one that hasn't breaks it into fragments. On our own text, one token carries 3.55 characters of English, 2.67 of Afaan Oromoo and 1.85 of Amharic.

If you are building for Ethiopian languages, measure with the tokenizer you will actually pay for. Published comparisons may be describing a different model.

Why "just paste everything in" doesn't work

With context windows of a million tokens, it is tempting to skip search entirely and paste the whole library into every prompt.

Our library now measures about 2.47 million tokens, so a one-million-token window holds well under half of it. It has also grown fast, from 3,937 chunks to 8,510 since we first published these figures.

Retrieval solves both problems. Across real questions in Amharic and Afaan Oromoo, Bini sends an average of 1,531 tokens per question, roughly 1,600 times less than the whole library. As the library grows, the cost per answer barely moves.

What we would tell other builders

• Ground answers in documents, not in model memory. For government procedures, fees and laws, being out of date is the most common way to be wrong.

• Show the source. People trust an answer they can check, especially when money or legal status depends on it.

• Measure your language's real cost. Tokenizers differ, and the difference can be more than threefold.

• Build the test set early. Our 138 questions tell us exactly where the gaps are.

Bini now speaks Amharic, English and Afaan Oromoo, and we are adding more Ethiopian languages. The full measurement, including how we count tokens and what we own versus license, is published at bina.et/amharic-ai.

You can try Bini free at bina.et/ai, in the browser or inside Telegram, with no download needed.

Ibrahim Kedir Bedru is the founder of BinaSmart (bina.et), based at JJ Darule Building in Addis Ababa.

Originally published on Medium.