Velqa documentation
Integrate Velqa like any OpenAI-compatible API, and learn every feature of the product.
Base configuration
base_url = "https://api.velqa.dev/v1"
api_key = "sk-..."
Getting started
What is Velqa — the OpenAI-compatible LLM API
Velqa.dev is an OpenAI- and Anthropic-compatible gateway for large language models, including Qwen, DeepSeek, GLM and MiniMax.
Get started with velqa.dev in 2 minutes
Velqa.dev is an OpenAI-compatible API gateway for open-weight models. Checkout is Stripe, in USD, by international card.
Velqa API authentication — `sk-...` key and Bearer header
, passed in the Authorization: Bearer HTTP header. That's the only secret the API needs: your account (email/OAuth) is only for the dashboard and billing.
Chat / LLM
Chat / LLM with Velqa — OpenAI-compatible API
Velqa's chat API follows the OpenAI POST /v1/chat/completions format exactly. key.
Streaming responses with Velqa — SSE, OpenAI-compatible
Velqa can return chat responses as a stream: tokens arrive as they're generated, via Server-Sent Events (SSE).
Tool calling / function calling with Velqa — OpenAI style
Let models call your functions with Velqa's tool calling: same format as OpenAI, with tools, tool_calls and role "tool" messages.
Velqa's Anthropic-compatible API — `/v1/messages` endpoint
Use Velqa's Anthropic-compatible /v1/messages endpoint so Claude Code and Anthropic SDKs run on open-weight models, with no Anthropic account.
Inference APIs
Embeddings with Velqa.dev
Velqa offers two multilingual embedding models (strong in French and Arabic), ideal for semantic search, RAG and document comparison.
Reranking with Velqa.dev — reorder your RAG results (FR / AR)
Reorder RAG results by real relevance with qwen3-reranker-8b on Velqa's /v1/rerank API, multilingual (French, Arabic), included in every plan.
Build a RAG pipeline on Velqa.dev — embeddings, pgvector and reranking
A RAG pipeline answers questions from your own documents. Velqa serves the three model calls — embedding, reranking, chat — behind a single API key.
Image generation (text-to-image) with Velqa.dev
dev/v1/images/generations, with three Black Forest Labs models covering drafts, everyday use and high quality.
Audio transcription (Speech-to-Text) with Velqa.dev
dev/v1/audio/transcriptions. Ideal for transcribing meetings, voice notes or customer calls.
Text-to-Speech with Velqa.dev
dev/v1/audio/speech. You send text and a voice, and the API returns the audio bytes directly (mp3).
Dashboard
The velqa.dev dashboard — manage your account and API keys
dev dashboard puts everything a developer needs in one place: create API keys, try models with no code, track your spend, and manage your balance.
velqa.dev API keys — create, limit, and revoke keys
Create velqa.dev API keys from the dashboard, restrict their models, set budget, RPM and TPM limits, and revoke a compromised key.
Track your velqa.dev usage — spend, history, and live quota
Read the velqa.dev Usage dashboard: monthly spend, daily breakdown by model, cost per project, request history and live quota.
velqa.dev Playground — try models with no API key or card
Try velqa.dev models in the browser with no code: publicly at /playground with a demo quota, or from your dashboard with no key to enter.
Billingbeta
velqa.dev plans & pricing — subscriptions and Recharge Boost
Velqa offers three monthly subscriptions (Starter, Dev, Pro) and a prepaid pay-as-you-go mode (Recharge Boost).
velqa.dev Recharge Boost — top up your USD balance via Stripe
Recharge Boost is Velqa's prepaid USD balance: pick an amount, pay by card through Stripe, and use every model with no subscription.
Automatic fallback to Boost on velqa.dev — stop getting blocked when your subscription runs out
Let spend spill over to your prepaid Boost balance when your subscription limits run out, instead of being blocked. Off by default.
Auto top-up and low-balance alerts on velqa.dev — never run out of credit
Auto top-up recharges your Boost balance by card when it drops below a threshold, and a low-balance email alert warns you in time.
Currency and billing on velqa.dev — USD
dev bills in US dollars (USD) through Stripe, by international card. The amount shown at Stripe Checkout is the amount charged.
Sandbox Agentbeta
Velqa Sandbox Agent — run code in an ephemeral Docker sandbox (BETA)
Velqa's Sandbox Agent lets you build AI agents that actually run code inside an isolated, ephemeral Docker sandbox, straight from the dashboard.
Create and configure a Sandbox agent (BETA)
An agent is a reusable profile you set up once and then start as many times as you need.
Sandbox Agent conversations and turns (BETA)
Once you've configured an agent, you talk to it through conversations made of turns.
Sandbox Agent tools (BETA)
Inside the sandbox, the agent doesn't just reply — it calls tools to act. During a turn, every tool call shows up live in the UI.
Sandbox Agent workspace and artifacts (BETA)
Every sandbox has a workspace: the folder where files live during a turn.
Page Agentbeta
Tool integrations
Tool integrations — connect your coding agents to Velqa
Any OpenAI-compatible tool (opencode, Cline, Roo Code, Kilo Code, Continue) works with Velqa: change the base_url and the key, no plugin.
Use Claude Code with MiniMax M3 or DeepSeek — velqa.dev
Configure Claude Code to use DeepSeek or MiniMax M3 through velqa.dev. Checkout is Stripe, in USD, by international card.
Use MiniMax M3 with opencode — velqa.dev
Configure opencode to use MiniMax M3 through velqa.dev. Checkout is Stripe, in USD, by international card.
Use MiniMax M3 with Cline — velqa.dev
Configure the Cline VS Code extension to use velqa.dev models. Checkout is Stripe, in USD, by international card.
Use MiniMax M3 with Roo Code — velqa.dev
Roo Code is a fork of Cline with advanced mode profiles (Code, Architect, Ask, Debug). It is configured the same way for an OpenAI-compatible provider.
Use MiniMax M3 with Kilo Code — velqa.dev
Kilo Code is an OpenAI-compatible VS Code coding agent. Configure it in a few fields to use velqa.dev; checkout is Stripe, in USD, by international card.
Use MiniMax M3 with Continue — velqa.dev
Continue is an open-source VS Code and JetBrains extension for chat, inline editing and agent workflows against the model of your choice.
Account
velqa.dev account — sign up, sign in, and verify your email
dev account takes seconds: email and password, or sign in with Google or GitHub.
velqa.dev settings — profile, language, ICE, and balance alerts
Manage your velqa.dev profile, dashboard language, billing address with ICE, low-balance alerts and account deletion from Settings.
velqa.dev referrals — earn $1 for every developer you invite
dev: your referred user gets $1 and you get $1 on their first payment. This guide explains how to grab your link and track your rewards from the dashboard.
Reference
Models available on velqa.dev
Every model available on velqa.dev, reachable through the OpenAI-compatible /v1/chat/completions and Anthropic-compatible /v1/messages APIs.
Rate limits and budgets on Velqa.dev
Velqa applies per-key API limits to protect your balance and smooth out load: a maximum throughput (RPM / TPM) per model, and a monthly USD budget per key.
Velqa.dev API error codes and how to fix them
dev/v1, their most common cause and the fix to apply. Every error response contains a JSON error object with an explicit message.
