MineAI O1 Pro & R3 Ry-Z• Deep Reasoning & 128k Context

The fastest way to build on MineAI Foundation Models

Prototype multimodal concepts in seconds, test prompts across 2 million tokens of native context, and export production-ready SDK code with real-time analytics.

2,097,152

Token Context Window

< 45 ms

Time to First Token (TTFT)

$5.00

Free Developer Credits

99.98%

Serverless Uptime SLA

Interactive Studio Preview

Test Real Prompts & Parameters Live

Switch between multimodal workloads and experience how MineAI models execute complex tasks.

mineai-studio-runtime: v2.5

Active Model:MineAI Vision FluxSub-50ms TTFT

Temp:0.7

Prompt InputToken Estimate: 48 tokens

Watch this 45-minute autonomous vehicle camera feed. Identify the exact frame and timestamp where the pedestrian enters the blind spot, and output the vehicle's speed and steering angle at that moment.

Need custom prompt inputs?Open in Studio Playground →

Synthesized OutputLatency: 210ms • Status: 200 OK

### Multimodal Temporal Synthesis - **Target Event Detected:** Pedestrian enters vehicle blind spot at **28:14.08** (Frame #50,824). - **Vehicle Telemetry at Detection:** - Speed: **34.2 km/h** (Decelerating at 1.8 m/s²) - Steering Angle: **-3.4°** (Micro-correction toward centerline) - Pedestrian Relative Velocity: **1.2 m/s** traversing from lateral sidewalk ```json { "timestamp": "00:28:14.08", "confidenceScore": 0.9984, "hazardLevel": "HIGH", "automatedActionTriggered": "EMERGENCY_PRE_BRAKE" } ``` Processed 50,824 video frames natively in unified token space without intermediate OCR or frame-drop artifacts.

Platform Capabilities

Architectural Superpowers Built for Scale

Everything you need to move from experimental concept to high-throughput production infrastructure.

2-Million-Token Infinite Context Window

Ingest 1 hour of raw 4K video, 60,000 lines of complex codebase, or 35 entire research books in a single prompt. No lossy vector chunking, no embedding drift, and 99.8% needle-in-a-haystack recall.

Loaded Multimodal Context1,842,500 / 2,097,152 Tokens (88%)

Video (45%)Codebase (25%)Audio (18%)12% Headroom Available

Sub-50ms TTFT

Engineered with speculative decoding and Flash kernels for real-time audio chat, interactive voice, and streaming agent feedback loops.

Legacy LLMs420 ms TTFT

MineAI R3 Ry-Z38 ms TTFT (11x Faster)

100% Guaranteed JSON Schema

Constrained decoding forces token generation into formal schema state machines. Zero syntax errors, zero missing brackets, zero hallucinations.

Schema Validation Rate:100.0% Guaranteed

Integrated Clerk Auth & Convex Realtime DB

Zero-setup backend infrastructure. User accounts, encrypted secret keys, telemetry logs, and inference credit balances are persistently synced and audited.

Table

users

Table

apiKeys

Table

prompts

Table

usageLogs

Model Lineup Matrix

MineAI Foundation Models

View Full Directory

MineAI O1 Pro

Recommended

mine:o1-paid

Standard

Thinking & Reasoning Workhorse

Our flagship multimodal model with deep internal reasoning, advanced coding capabilities, and complex multi-step document synthesis.

Context Window:128,000

Pricing / 1M In:$5.00

Pricing / 1M Out:$15.00

Try in Studio Playground

MineAI R3 Ry-Z

New

mine:r3-ry-z

Fast

High-Performance Autonomous Reasoning

Optimized for automated programming, debugging, full-stack refactoring, and structured reasoning pipelines.

Context Window:128,000

Pricing / 1M In:$3.50

Pricing / 1M Out:$9.00

Try in Studio Playground

MineAI R3 Ry-Y

mine:r3-ry-y

Fast

Balanced General Intelligence

A fast, balanced intelligence model suitable for natural conversation, text transformation, summarization, and workflow automation.

Context Window:64,000

Pricing / 1M In:$2.50

Pricing / 1M Out:$6.00

Try in Studio Playground

Developer SDK

Integrate in 3 Lines of Code

Connect using the MineAI SDK for Python, Node.js, and Go, or invoke the standard REST API directly using your secret key.

Drop-in replacement for OpenAI

Streaming token generation support

Automatic retries and backoff handling

Get Live Credentials

# Install: pip install mineai-sdk
from mineai import MineAI

client = MineAI(api_key="mine_live_...")

response = client.chat.completions.create(
    model="mine:o1-paid",
    messages=[{"role": "user", "content": "Synthesize architectural insights across 128k tokens"}],
    temperature=0.7,
    max_tokens=4096,
)

print(response.choices[0].message.content)

Got Questions?

Frequently Asked Questions

Everything about model capabilities, billing credits, and developer tools.

Start Free with $5 Credits

Ready to build the future of multimodal AI?

Launch Studio Playground →Browse Model Specs

# Install: pip install mineai-sdk from mineai import MineAI client = MineAI(api_key="mine_live_...") response = client.chat.completions.create( model="mine:o1-paid", messages=[{"role": "user", "content": "Synthesize architectural insights across 128k tokens"}], temperature=0.7, max_tokens=4096, ) print(response.choices[0].message.content)