Start building

One interface to mainstream language models.

Better prices, steadier calls, and no subscription. Every model draws on the same credit.

Get API keyBrowse models
Gateway · /v1/chat/completionsUSD / 1M tokens
ModelInputOutput
Claude Opus 4.8Anthropic$5.00$25.00
GPT-5.5OpenAI$5.00$30.00
DeepSeek V4 ProDeepSeek$1.32$3.96
GPT-5.6 LunaOpenAI$0.20$1.20
One key · one credit · billed per tokenOpenAI-compatible
Available models
100+
Model providers
20+
API key for every model
1
Platform subscription
$0

Every model behind one interface.

Call text, image, video, and audio models through the same request, and keep the controls a team needs in production.

  • Text, image, video, and audio

    A unified interface to mainstream model capabilities.

    Browse all models

  • Higher availability

    When a provider is down, calls switch to another available line.

    Read the gateway reference

  • Cost vs. performance

    Balance cost, speed, and quality for each task.

    Compare model prices

  • Custom data policy

    Send requests only to the models and providers your organization allows.

    Ask about data policy

100+ models from 20+ providers.

Find models by use, trend, and fit for the task.

View all models
  • AnthropicDeep reasoning

    Claude Opus 4.8

    Claude Opus 4.8 is Anthropic's most capable generally available model in the Opus family. It supports text, image, and file inputs with text output, reasoning support, and a 1M-token context window.

    Context
    1M
    Input
    $5.00
    Output
    $25.00
    Use this model
  • OpenAIAgent work

    GPT-5.5

    GPT-5.5 is OpenAI’s frontier model designed for complex professional workloads, with stronger reasoning, higher reliability, and improved token efficiency on hard tasks.

    Context
    1.1M
    Input
    $5.00
    Output
    $30.00
    Use this model
  • DeepSeekLong context

    DeepSeek V4 Pro

    DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window.

    Context
    1M
    Input
    $1.32
    Output
    $3.96
    Use this model

Pay as you go

No subscription. Every model shares the same credit.

Free credits
$1
Platform subscription
$0
Billing
Per token
Get API keyView full pricing

Compare models side by side on the same task.

Check the answer, the latency, and the cost first. Then copy the API request, or add the winning model to your Agent.

Model labCompare mode

  • Claude Opus 4.81.8s · $0.06

    The market reads best through three structural shifts: model capability is becoming a commodity, tool calling is standardizing, and products are moving from answering questions to delivering work.

    8 sources citedAdd to Agent

  • GPT-5.51.2s · $0.04

    Competition among Agent platforms is moving from model access to reliable execution, connections to external systems, and the ability to keep running in production.

    6 sources citedAdd to Agent

Analyze the Agent platform market and return the key conclusions with sources.Send

Copy API requestSave model set

Start calling models in three steps.

  1. 01

    Create an account

    Sign up and SandBase creates your personal workspace. Invite your team when you need to.

  2. 02

    Add credit

    Top up once. The same credit works across every model and provider.

  3. 03

    Get an API key

    Create a key and copy your first OpenAI-compatible request.

First request

curl https://api.sandbase.ai/v1/chat/completions \
  -H "Authorization: Bearer $SANDBASE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"anthropic/claude-opus-4.8","messages":[{"role":"user","content":"Hello"}]}'

Model chosen? Put it to work on a real task.

Bring SandBase into Claude, ChatGPT, or Codex to prove the value first, then turn the work into a complete Agent.

Build an AgentExplore API, SDK & MCP