MoryFlow
Settings

AI Model Configuration

MoryFlow supports 20+ AI providers—pick the one that works best for you.

AI Model Configuration

MoryFlow supports multiple AI providers. Choose the combination that fits your workflow, budget, and privacy requirements.

Supported Providers

Direct Providers

ProviderFeaturesRecommended Models
OpenAIMost popular, rich ecosystemGPT-4o, GPT-4 Turbo
AnthropicClaude series, great with long textClaude 3.5 Sonnet
GoogleGemini seriesGemini Pro
xAIGrok seriesGrok-2
DeepSeekAffordable, multilingualDeepSeek V3
MoonshotChinese optimizedMoonshot-v1
Zhipu AIChinese company, strong ChineseGLM-4

Aggregated Providers

ProviderFeatures
OpenRouterOne key for multiple models, great for beginners
AIHubMixChinese aggregation service

Local Models

MethodFeatures
OllamaOpen source, supports many models, fully offline

Configuring API Keys

How to Configure

  1. Open Settings → Providers
  2. Find the provider you want
  3. Click the configure button
  4. Enter your API Key
  5. Save

Where to Get API Keys

  1. Go to openrouter.ai
  2. Sign up
  3. Go to Dashboard → Keys
  4. Create a new API Key

OpenAI

  1. Go to platform.openai.com
  2. Sign in
  3. Go to API Keys page
  4. Create new Secret Key

Anthropic

  1. Go to console.anthropic.com
  2. Sign up/in
  3. Go to API Keys
  4. Create new key

DeepSeek

  1. Go to platform.deepseek.com
  2. Sign up
  3. Go to API Keys
  4. Create new key

Selecting Models

After configuring an API Key, you can select models from the top of the chat panel:

Model Recommendations

Everyday Use

NeedRecommended ModelReason
All-aroundClaude 3.5 SonnetGreat quality, reasonable price
Fast responsesGPT-4o-miniQuick, affordable
Chinese writingDeepSeek V3Strong Chinese, cost-effective

Professional Scenarios

ScenarioRecommended ModelReason
CodingClaude 3.5 SonnetBest code capabilities
Long documentsClaude 3 Opus200K context
Privacy sensitiveOllama local modelsFully offline

Budget-Conscious

BudgetRecommended Approach
Just tryingOpenRouter free credits
Low costDeepSeek V3 (~$0.07/10K words)
Completely freeOllama local models

Adding Custom Models

If your model isn't in the presets, you can add it:

  1. Open Settings → Providers
  2. Click "Add Custom Model"
  3. Fill in info:
    • Model name
    • Model ID
    • Provider
    • Context window size

Model Parameters

For those who like to tinker:

ParameterDescriptionSuggested Value
TemperatureControls randomness0.7 (default)
Max TokensMaximum output lengthAdjust as needed
Top PNucleus sampling1 (default)

Tip

Default parameters usually work fine. Lower temperature if responses feel too random; raise it if too rigid.

Common Issues

API Key Invalid

  1. Check you copied it completely (no extra spaces)
  2. Confirm key hasn't expired
  3. Confirm account balance is sufficient

Model Responds Slowly

  1. Switch to a faster model (e.g., GPT-4o-mini)
  2. Check your network
  3. Consider local models

Do I Need a VPN?

  • OpenAI, Anthropic, Google: May need depending on region
  • DeepSeek, Zhipu, Moonshot: Generally accessible everywhere
  • OpenRouter: Varies by region

On this page

AI Model ConfigurationSupported ProvidersDirect ProvidersAggregated ProvidersLocal ModelsConfiguring API KeysHow to ConfigureWhere to Get API KeysOpenRouter (Recommended for Beginners)OpenAIAnthropicDeepSeekSelecting ModelsModel RecommendationsEveryday UseProfessional ScenariosBudget-ConsciousAdding Custom ModelsModel ParametersCommon IssuesAPI Key InvalidModel Responds SlowlyDo I Need a VPN?