CoroidDark Factory

Model Selection

Choose your model profile

Coroid uses a blend of US and Chinese models, each selected for their strengths. Pick the profile that fits your task and budget.

Start Free
A plot of the six model profiles, positioned by output cost per million tokens along the horizontal axis and by quality along the vertical axis. Each point is labelled with its profile name; the EU-only profile is drawn as an open ring. The same profiles are listed in full underneath, where each can be expanded for its models, pricing, and recommended uses.

Cost against quality

Select a profile to open it

GoodExcellentOutstanding
$0.00$1.00$5.00$10.00$20.00

Output cost per million tokens

EU-hosted models only

Larger mark = faster

  1. Zero CostFree

    A no-cost profile where Coroid routes each task to appropriate cost-efficient models. Ideal for trying Coroid, lightweight work, and basic testing while model choices may change over time.

    Models

    Nemotron 3 Super

    Best For

    Evaluating Coroid · Personal projects · Lightweight coding tasks · Zero-budget experimentation

    Input
    $0.00 / 1M tokens
    Output
    $0.00 / 1M tokens
  2. Smart & AffordableBudget

    Cost-effective models that deliver excellent value. Perfect for routine tasks, documentation, and straightforward implementations.

    Models

    GLM-5.2 · Nemotron 3 Nano 30B · Qwen3.7 Flash · DeepSeek V4 Flash 0731 · GPT-5.6 Luna

    Best For

    Simple feature implementations · Documentation and comments · Code formatting and linting · Refactoring tasks

    Input
    $0.03 - $1.00 / 1M tokens
    Output
    $0.13 - $6.00 / 1M tokens
  3. Maximum IntelligenceSmartest

    State-of-the-art models for complex architectural decisions, debugging, and critical code paths.

    Models

    Fable 5 · Sonnet 5 · GPT-5.6 Terra · Opus 4.8

    Best For

    Complex architecture design · Debugging intricate issues · Security-critical code · Algorithm optimisation

    Input
    $2.00 - $10.00 / 1M tokens
    Output
    $10.00 - $50.00 / 1M tokens
  4. Speed FirstQuick

    Lightning-fast responses for iterative development, quick fixes, and when you need answers now.

    Models

    GPT-5.6 Terra · Nemotron 3 Nano 30B · Grok Fast · Google Gemini 3.1 Flash Lite · DeepSeek V4 Flash 0731 · GPT-5.4 Nano

    Best For

    Rapid prototyping · Quick iterations · Simple fixes · Exploration tasks

    Input
    $0.06 - $2.50 / 1M tokens
    Output
    $0.20 - $15.00 / 1M tokens
  5. Maximum SpeedLudicrous

    Fastest available models with relaxed quality gates. For when velocity matters most.

    Models

    DeepSeek V4 Flash 0731 · Nemotron 3 Nano 30B · Qwen3.7 Flash · GPT-5.6 Luna

    Best For

    Emergency fixes · Maximum velocity · Draft implementations · Rapid experimentation

    Input
    $0.03 - $1.00 / 1M tokens
    Output
    $0.13 - $6.00 / 1M tokens
  6. GDPR CompliantEU Only

    GDPR-compliant AI profile using models hosted exclusively in Europe. All data processing stays within EU jurisdiction with full Schrems II protection.

    Models

    GLM-5.1 · MiniMax M2.7 · Devstral 2 · GLM-5.2

    Best For

    GDPR-sensitive projects · EU data residency requirements · Privacy-critical applications · Enterprise compliance

    Input
    $0.30 - $1.00 / 1M tokens
    Output
    $1.20 - $3.20 / 1M tokens

Global Model Selection

We believe in using the best tools for the job, regardless of origin. Our platform intelligently routes your tasks to the most appropriate models:

US Models: Best for complex reasoning, security-critical code, and when you need the absolute highest quality.

Chinese Models: Excellent value for routine tasks, strong coding capabilities, and cost-effective scaling.