5-Minute ByteSpike integration — from DOSIA download to first chat

Five steps, ~5 minutes: install DOSIA, sign up for ByteSpike, mint an API key, wire the provider into DOSIA Settings, and send your first prompt. By the end you'll be routing through one key to Claude / GPT / Gemini / DeepSeek and watching the cost land at sub-fen.

May 14, 2026ByteSpike Team5 min read

If you already have DOSIA and a ByteSpike account, skip to Step 4. Otherwise, here is the most boring path from zero to a working multi-model setup.

Step 1 — Download DOSIA

DOSIA is a native macOS client that hosts your multi-provider agent workflows. Grab the universal installer at bytespike.ai/dosia — same DMG handles both Apple Silicon and Intel. First launch asks for the usual accessibility / file-system permissions; grant them and you land on an empty conversation.

DOSIA download page at bytespike.ai/dosia showing universal macOS installer link

Step 2 — Sign up for ByteSpike

Head to console.bytespike.ai/sign-up. Email + a six-digit code is the entire flow — no card capture, no team setup wizard. About 30 seconds total. New accounts come with a small trial credit so you can verify routing end-to-end before adding billing.

Step 3 — Mint your first API key

Once signed in, go to console.bytespike.ai/keys and click Create key. Name it something you'll recognize ("DOSIA — local" works), copy the sk-byts-… string, and store it where you keep secrets. The console will not show it again. Verify it's live with the balance endpoint before you paste it into DOSIA:

ByteSpike console /keys page showing the Create key dialog with name field and one-time secret reveal

bash

# Confirm the key is live (sub-second round trip)
curl https://llm.bytespike.ai/v1/balance \
  -H "Authorization: Bearer sk-byts-..."

# expected response shape
# { "balance": "0.1000", "currency": "USD", "estimated_calls_remaining": 4500 }

Step 4 — Add ByteSpike as a provider in DOSIA

Open DOSIA → Settings → Providers & Models → Add provider → pick ByteSpike. Three fields: • baseUrl: https://llm.bytespike.ai/v1 • apiKey: the sk-byts-… string from Step 3 • default model: deepseek-v4-flash (cheap, fast, multilingual) DOSIA pulls the live model list from ByteSpike's /v1/models endpoint, so once the key is valid you'll see all 23 models populate the picker.

DOSIA Settings → Providers & Models → Add ByteSpike provider dialog with baseUrl, apiKey, and default model fields filled in

Step 5 — Your first chat

Back on the main DOSIA screen, send "hi" — or any prompt. You'll watch DeepSeek V4 Flash stream a response in under a second. Open console.bytespike.ai/usage in a browser tab and you'll see the request land with an estimated cost around ¥0.0001 — the receipt arrives within a few seconds of the response finishing.

DOSIA chat screen showing a first hi prompt with DeepSeek V4 Flash response, alongside a console /usage tab showing the same request logged at ¥0.0001

That's the whole loop. From here you can: switch DOSIA's model picker to Claude or GPT for the same key; route image / video / audio generation through the same ByteSpike key (multimodal endpoints are first-class); install role files for engineering / sales / ops; or wire DOSIA into your Lark / Slack / GitHub via the skill marketplace.

Download DOSIA at bytespike.ai/dosia. Open a ByteSpike account at console.bytespike.ai/sign-up. Both are free to start. Tell us what's friction-y at hello@bytespike.ai — onboarding is a moving target and we read every reply.