DR.PANDA AI PLATFORM AI infrastructure for private deployments Explore the platform
Dr.PandaAI PLATFORM
● DR.PANDA MODEL & API PLATFORM

AI APIs built
for real work.

A private AI platform for teams running models in their own environment. Connect language, vision, image and voice services through one developer API layer.

Get started ↗View API examples →
✓ Explore Dr.Panda’s models, APIs and tools for building with AI.
PLATFORM OVERVIEWOne API. Multiple capabilities.
DP / 01
API
Dr.Panda API GatewayOpenAI + Anthropic compatible
ROUTING
01LanguageCore · ReasoningCHAT / VISION / TOOLS
02ImageGenerate · EditPICTURE API
03VoiceTranscribe · Process · SpeakAUDIO SERVICES
API requests use your keyUsage metered by tokens
OPENAI-COMPATIBLEANTHROPIC-COMPATIBLECHAT & VISIONIMAGE GENERATION & EDITINGVOICE SERVICES
PLATFORM HIGHLIGHTS

One surface for
more useful work.

View platform notes →
LANGUAGE & VISION

Familiar model APIs

OpenAI-compatible chat completions and Anthropic-compatible messages support documented chat, tool and visual-input paths.

IMAGE & VOICE

Multimodal building blocks

Explore documented image generation/editing and speech recognition, text polishing and voice output service directions.

CONTEXT & USAGE

Plan larger requests

See the existing client default and explicitly negotiated context limits, then estimate request size locally.

MODEL TIERS · CONCEPT NAMES

One model layer.
Clear choices.

See API patterns →

Choose a model tier to match your workload. All examples on this page are illustrative; confirm available model IDs before integrating.

CONCEPT TIER↗
GO

Dr.Panda Core

A balanced model tier for everyday chat, structured tasks, visual input and tool-assisted workflows.

Chat completionsVision inputTool calls
drpanda-core*OpenAI-compatible
CONCEPT TIER↗
Pro

Dr.Panda Reasoning

A higher-capability tier for deeper analysis, longer tasks and more demanding application flows.

Chat completionsVision inputTool calls
drpanda-reasoning*Anthropic-compatible too
IMAGE SERVICE↗
IMG

Picture API

Generate images from prompts and edit existing images through the gateway.

Image generationImage editingLegacy aliases supported
imagesOne routed service

*Illustrative Dr.Panda model IDs. Confirm currently enabled IDs and pricing before production use.

PLATFORM CAPABILITIES

Beyond chat, one gateway.

A practical set of AI building blocks, organized around what your product needs to do.

01 / LANGUAGE

Chat, tools & vision

OpenAI-compatible chat completions and Anthropic-compatible messages, with tool calls and image input.

/v1/chat/completions/v1/messages
02 / IMAGE

Generate & edit

Image generation and editing are routed through the Picture service, including supported legacy aliases.

Image generationImage editing
03 / VOICE

Speech in & out

The wider service stack includes speech recognition, transcript polishing, voice output and live audio workflows with flexible fallback.

ASR · Polish · TTSReal-time audio · HTTP fallback
i

Long context, handled carefully. Existing clients default to 262,144 tokens. The gateway can negotiate up to 327,680 when a client explicitly requests the supported capacity.

262,144 DEFAULT → 327,680 NEGOTIATED
DR.PANDA PRODUCT MATRIX

Choose a starting
point for your app.

MODEL API

Dr.Panda API

Connect through familiar chat-completions and messages request formats.

Read API docs ↗MODEL CATALOGUE

Models

Compare model families and capabilities for your workload.

Explore models ↗AI PLAYGROUND

Dr.Panda Chat

Explore practical prompts across models, tools and multimodal workflows.

Explore the playground ↗MEDIA APIS

Picture & Voice

Explore documented image and speech-service directions.

View capabilities ↗
A DEVELOPER WORKFLOW

Connect the pieces
you already use.

Start with an API shape your application understands, select a confirmed model ID, then add tools or supported media as your workflow needs them.

Explore integration patterns ↗
01Choose a modelUse an ID confirmed for your environment.
02Send a requestUse chat completions or messages.
03Add capabilitiesTools, image or voice paths where supported.
04Track usageEstimate locally; confirm actual billing rates.
A CONSISTENT DEVELOPER INTERFACE

Two API styles.
One gateway.

Connect through familiar OpenAI-compatible chat completions or Anthropic-compatible messages. Route supported models through one consistent integration layer.

See integration notes ↗
01OpenAI-compatibleChat completions, vision and tool calls.
02Anthropic-compatibleMessages API for existing clients.
03Context negotiationPreserve older defaults; opt into supported capacity.
chat-completions.js
// OpenAI-compatible request
const response = await fetch(
  "https://YOUR_API_BASE_URL/v1/chat/completions", {
    method: "POST",
    headers: {
      "Authorization": `Bearer ${process.env.DR_PANDA_API_KEY}`,
      "Content-Type": "application/json"
    },
    body: JSON.stringify({
      model: "drpanda-core",
      messages: [{ role: "user", content: "Hello, Dr.Panda." }]
    })
  }
);
const result = await response.json();
KEEP API KEYS IN YOUR SERVER ENVIRONMENTJavaScript
API QUICKSTART · 01–04

Connect your app
to Dr.Panda.

Confirm your deployment address and model, keep the key on your server, then send a request using a familiar API format.

  1. 01

    Confirm your endpoint

    Use the API base URL assigned to your deployment.

  2. 02

    Choose an API format

    Chat Completions /v1/chat/completions or Messages /v1/messages.

  3. 03

    Set model and key

    Use a confirmed model ID. Keep your API key in a server environment variable.

  4. 04

    Send your request

    Call the endpoint from your application server and handle the response.

Read the API setup guide →
OPENAI-COMPATIBLE · CHAT COMPLETIONSSERVER-SIDE REQUEST
POST/v1/chat/completions
curl -X POST "${DR_PANDA_API_BASE_URL}/v1/chat/completions" \
  -H "Authorization: Bearer ${DR_PANDA_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "YOUR_CONFIRMED_MODEL_ID",
    "messages": [{"role": "user", "content": "Hello, Dr.Panda."}]
  }'
Use the base URL and model ID provided for your deployment.API reference ↗
AI PLAYGROUND

Explore the request flow.

Explore common AI workflows through a simple interactive playground.

CHAT COMPLETION · PLAYGROUNDDR.PANDA AI WORKSPACE
DP
DR.PANDA · AI ASSISTANT

An API gateway gives your application one dependable entry point, then routes each request to the right model and service.

Explore model capabilities and API integration patterns.A practical place to explore AI workflows
TOKEN USAGE

Know what your
requests consume.

Track input and output separately. Use the estimate below to understand request size before sending it to a model.

COST ESTIMATOR
TOKEN ESTIMATORUSAGE ESTIMATE
INPUT TOKENS · APPROX.25
OUTPUT BUDGET
1,024
DEFAULT CONTEXT262,144 tokens
NEGOTIATED CAPACITYUp to 327,680 tokens

Token counts here are rough estimates based on text length. Actual tokenization and charges depend on the model and account rate.

DEVELOPER RESOURCES

Integrate with
confidence.

Start with familiar API shapes, then connect the model and media capabilities your application needs.

Read integration notes ↗
01Chat completionsOpenAI-compatible request shape↗02Messages APIAnthropic-compatible requests↗03Vision & tool callsMultimodal requests and tools↗04Images & usageGeneration, editing and token estimates↗
IMPLEMENTATION NOTES

Common questions.

Key details for connecting your application to the documented services.

What can I build with Dr.Panda?+

Connect chat, tool use, image and voice capabilities through API patterns that fit your application.

Which API formats are supported?+

Use familiar OpenAI-compatible chat completions or Anthropic-compatible messages, with tool calls and image input.

How does long context work?+

Existing clients keep a 262,144-token default. Clients that request the supported capacity can negotiate up to 327,680 tokens; higher requests are clamped to the supported maximum.

Which media services are documented?+

Explore image generation and editing, speech recognition, transcript polishing and voice workflows.

ABOUT DR.PANDA

Useful intelligence.
Clear interfaces.

Dr.Panda is designed for teams running AI in their own environment. One developer surface connects model access and multimodal tools to the products they already build.

Explore the platform without an account. When you are ready to make live requests, call the API from your own server and keep your credentials in your environment.

Explore developer resources →
01
ONE API · MULTIPLE CAPABILITIES

Build with
Dr.Panda.

Get started ↗