Model Provider List

OpenCode supports 75+ model providers via AI SDK and Models.dev

📝 Course Notes

Key takeaways from this lesson:

Model Provider Notes

Configuration

Adding a provider requires two steps:

Use /connect command to add API Key (stored in ~/.local/share/opencode/auth.json)
Configure provider in opencode.json (optional, for custom options)

OpenCode Zen (Recommended for Beginners)

Official OpenCode verified model list, ready to use.

bash

# 1. Run in TUI
/connect

# 2. Select opencode, visit opencode.ai/auth to get API Key

# 3. View available models
/models

Get API Key: opencode.ai/auth

International Models

Anthropic Claude

Model	Description
`claude-sonnet-4-20250514`	Latest balanced (recommended)
`claude-opus-4-20250514`	Most powerful
`claude-3-5-haiku-20241022`	Fast model

Configuration:

bash

/connect  # Select Anthropic

# Options:
# - Claude Pro/Max (browser auth)
# - Create an API Key (create new key)
# - Manually enter API Key

Get API Key: console.anthropic.com

OpenAI

Model	Description
`gpt-4o`	Flagship multimodal
`gpt-4o-mini`	Budget option
`o1`	Reasoning model
`o3-mini`	Latest reasoning

Configuration:

bash

/connect  # Search OpenAI

Get API Key: platform.openai.com

Google Gemini

Via Vertex AI.

Model	Description
`gemini-2.0-flash`	Latest fast version
`gemini-2.0-pro`	Professional
`gemini-1.5-pro`	Long context

Configuration:

bash

# Set Google Cloud project ID (required)
export GOOGLE_CLOUD_PROJECT=your-project-id
# Or use GCP_PROJECT / GCLOUD_PROJECT

# Set region (optional, default us-east5)
export VERTEX_LOCATION=us-east5
# Or use GOOGLE_CLOUD_LOCATION

Google Vertex AI requires authentication via gcloud auth application-default login or service account. OpenCode automatically uses Application Default Credentials.

xAI Grok

Model	Description
`grok-2`	Latest version
`grok-2-mini`	Budget option

Configuration:

bash

/connect  # Search xAI

Get API Key: console.x.ai

Mistral

Open-source model leader, supports Mistral Large, Codestral, etc.

Model	Description
`mistral-large-latest`	Most capable
`mistral-small-latest`	Fast response
`codestral-latest`	Code optimized

bash

/connect  # Search Mistral

Get API Key: console.mistral.ai

Cohere

Enterprise NLP capabilities, supports Rerank, Embed.

bash

/connect  # Search Cohere

Get API Key: dashboard.cohere.com

Perplexity

Integrated search capabilities, real-time information.

bash

/connect  # Search Perplexity

Get API Key: perplexity.ai/settings/api

API Key format: pplx-...

GitHub Copilot

Use Copilot subscription.

bash

/connect  # Search GitHub Copilot
# Visit github.com/login/device to enter authorization code

Some models require Pro+ subscription, certain models need to be manually enabled in GitHub Copilot settings.

DeepSeek

Excellent value, works globally. Also popular in China for direct local access.

Model	Description
`deepseek-chat`	General chat
`deepseek-reasoner`	Reasoning model (R1)

Configuration:

bash

# 1. Run /connect, search DeepSeek
/connect

# 2. Enter API Key

# 3. Select model
/models

Get API Key: platform.deepseek.com

Chinese Models

Moonshot AI

Kimi K2 model.

Model	Description
`kimi-k2`	Latest model

Configuration:

bash

/connect  # Search Moonshot AI

Get API Key: platform.moonshot.ai

MiniMax

Model	Description
`M2.1`	Latest model

Configuration:

bash

/connect  # Search MiniMax

Get API Key: platform.minimax.io

Z.AI (Zhipu)

GLM series models.

Model	Description
`GLM-4.7`	Latest model

Configuration:

bash

/connect  # Search Z.AI
# If subscribed to GLM Coding Plan, select Z.AI Coding Plan

Get API Key: z.ai

Cloud Platforms

Amazon Bedrock

bash

# Environment variable
AWS_PROFILE=my-profile opencode

# Or config file

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "amazon-bedrock": {
      "options": {
        "region": "us-east-1",
        "profile": "my-aws-profile"
      }
    }
  }
}

Azure OpenAI

bash

/connect  # Search Azure OpenAI

If you encounter "I'm sorry, but I cannot assist" error, change content filter from DefaultV2 to Default.
Azure OpenAI is configured via /connect, credentials are automatically stored.

Azure Cognitive Services

bash

/connect  # Search Azure Cognitive Services

# Set resource name
export AZURE_COGNITIVE_SERVICES_RESOURCE_NAME=your-resource-name

Cloudflare Workers AI

Cloudflare edge network, global low latency.

bash

/connect  # Search Cloudflare Workers AI

# Or set environment variables
export CLOUDFLARE_API_KEY=your-api-token
export CLOUDFLARE_ACCOUNT_ID=your-account-id

Get API Token: dash.cloudflare.com → My Profile → API Tokens

GitLab

GitLab Duo Chat, deeply integrated with GitLab.

bash

/connect  # Search GitLab

# For enterprise instances
export GITLAB_INSTANCE_URL=https://gitlab.company.com

Get Token: gitlab.com → Settings → Access Tokens

Token format: glpat-...

Aggregation Platforms

OpenRouter

Access 100+ models with one API Key.

bash

/connect  # Search OpenRouter

Custom Models:

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "openrouter": {
      "models": {
        "moonshotai/kimi-k2": {
          "options": {
            "provider": {
              "order": ["baseten"],
              "allow_fallbacks": false
            }
          }
        }
      }
    }
  }
}

Get API Key: openrouter.ai

Groq

Ultra-fast inference.

bash

/connect  # Search Groq

Get API Key: console.groq.com

Cerebras

Ultra-fast inference, supports Qwen3 Coder 480B.

bash

/connect  # Search Cerebras

Get API Key: inference.cerebras.ai

Fireworks AI

bash

/connect  # Search Fireworks AI

Get API Key: app.fireworks.ai

Deep Infra

bash

/connect  # Search Deep Infra

Get API Key: deepinfra.com/dash

Together AI

bash

/connect  # Search Together AI

Get API Key: api.together.ai

Hugging Face

Access open-source models from 17+ providers.

bash

/connect  # Search Hugging Face

Get Token: huggingface.co/settings/tokens

Baseten

bash

/connect  # Search Baseten

Get API Key: app.baseten.co

Cortecs

Supports Kimi K2 Instruct.

bash

/connect  # Search Cortecs

Get API Key: cortecs.ai

Nebius Token Factory

bash

/connect  # Search Nebius Token Factory

Get API Key: tokenfactory.nebius.com

IO.NET

Provides 17+ models.

bash

/connect  # Search IO.NET

Get API Key: ai.io.net

Venice AI

bash

/connect  # Search Venice AI

Get API Key: venice.ai

OVHcloud AI Endpoints

bash

/connect  # Search OVHcloud AI Endpoints

Get API Key: ovh.com/manager → Public Cloud → AI & Machine Learning → AI Endpoints

SAP AI Core

Access 40+ models (OpenAI, Anthropic, Google, Amazon, Meta, etc.).

bash

/connect  # Search SAP AI Core

Requires Service Key JSON (containing clientid, clientsecret, url, serviceurls.AI_API_URL).

Cloudflare AI Gateway

Unified access to multiple providers via Cloudflare, with unified billing.

bash

# Set environment variables
export CLOUDFLARE_ACCOUNT_ID=your-account-id
export CLOUDFLARE_GATEWAY_ID=your-gateway-id

/connect  # Search Cloudflare AI Gateway

Vercel AI Gateway

Unified access to multiple providers via Vercel, at-cost pricing with no markup.

bash

/connect  # Search Vercel AI Gateway

Configure Routing Order:

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "vercel": {
      "models": {
        "anthropic/claude-sonnet-4": {
          "options": {
            "order": ["anthropic", "vertex"]
          }
        }
      }
    }
  }
}

Helicone

LLM observability platform, provides logging, monitoring and analytics.

bash

/connect  # Search Helicone

Get API Key: helicone.ai

Custom Headers:

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "helicone": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Helicone",
      "options": {
        "baseURL": "https://ai-gateway.helicone.ai",
        "headers": {
          "Helicone-Cache-Enabled": "true",
          "Helicone-User-Id": "opencode"
        }
      }
    }
  }
}

ZenMux

bash

/connect  # Search ZenMux

Get API Key: zenmux.ai/settings/keys

Ollama Cloud

Cloud-based Ollama service.

bash

/connect  # Search Ollama Cloud

Pull model info locally first: ollama pull gpt-oss:20b-cloud

Get API Key: ollama.com → Settings → Keys

Local Models

Ollama

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "ollama": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Ollama (local)",
      "options": {
        "baseURL": "http://localhost:11434/v1"
      },
      "models": {
        "llama3.1": {
          "name": "Llama 3.1"
        }
      }
    }
  }
}

If tool calling doesn't work, try increasing Ollama's num_ctx, recommended 16k-32k.

Install: ollama.ai

LM Studio

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "lmstudio": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "LM Studio (local)",
      "options": {
        "baseURL": "http://127.0.0.1:1234/v1"
      },
      "models": {
        "google/gemma-3n-e4b": {
          "name": "Gemma 3n-e4b (local)"
        }
      }
    }
  }
}

Install: lmstudio.ai

llama.cpp

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "llama.cpp": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "llama-server (local)",
      "options": {
        "baseURL": "http://127.0.0.1:8080/v1"
      },
      "models": {
        "qwen3-coder:a3b": {
          "name": "Qwen3-Coder: a3b-30b (local)",
          "limit": {
            "context": 128000,
            "output": 65536
          }
        }
      }
    }
  }
}

Custom Providers

Add any OpenAI-compatible provider:

bash

# 1. Run /connect, select Other
/connect

# 2. Enter provider ID (e.g., myprovider)

# 3. Enter API Key

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "myprovider": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "My Provider",
      "options": {
        "baseURL": "https://api.myprovider.com/v1"
      },
      "models": {
        "my-model": {
          "name": "My Model",
          "limit": {
            "context": 200000,
            "output": 65536
          }
        }
      }
    }
  }
}

Configuration Options:

npm - AI SDK package name, use @ai-sdk/openai-compatible for OpenAI-compatible
name - UI display name
options.baseURL - API endpoint
options.apiKey - API Key (optional, set when not using auth)
options.headers - Custom request headers
models - Available models list
limit.context - Max input tokens
limit.output - Max output tokens

Custom Base URL

Set custom endpoints for any provider (e.g., proxy services):

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "anthropic": {
      "options": {
        "baseURL": "https://my-proxy.com/v1"
      }
    }
  }
}

Model Selection Guide

Requirement	Recommendation	Reason
Easiest to start	Claude Sonnet 4	Best coding, reliable
Best reasoning	Claude Opus 4	Industry leading
Best value	DeepSeek	Affordable, capable
Best coding	Claude Sonnet 4	Professional programming
Long documents	Gemini 1.5 Pro	Ultra-long context
Fully offline	Ollama + Llama3.1	Local runtime
Multi-model access	OpenRouter	One key for all

Troubleshooting

Check Authentication: Run opencode auth list to view configured credentials
Custom Provider Issues:
- Ensure provider ID in /connect matches config file
- Use correct npm package (e.g., @ai-sdk/openai-compatible)
- Check options.baseURL is correct

Connect Models - Configuration tutorial
Config Reference - Config file details

Model Provider List ​

📝 Course Notes ​

Configuration ​

OpenCode Zen (Recommended for Beginners) ​

International Models ​

Anthropic Claude ​

OpenAI ​

Google Gemini ​

xAI Grok ​

Mistral ​

Cohere ​

Perplexity ​

GitHub Copilot ​

DeepSeek ​

Chinese Models ​

Moonshot AI ​

MiniMax ​

Z.AI (Zhipu) ​

Cloud Platforms ​

Amazon Bedrock ​

Azure OpenAI ​

Azure Cognitive Services ​

Cloudflare Workers AI ​

GitLab ​

Aggregation Platforms ​

OpenRouter ​

Groq ​

Cerebras ​

Fireworks AI ​

Deep Infra ​

Together AI ​

Hugging Face ​

Baseten ​

Cortecs ​

Nebius Token Factory ​

IO.NET ​

Venice AI ​

OVHcloud AI Endpoints ​

SAP AI Core ​

Cloudflare AI Gateway ​

Vercel AI Gateway ​

Helicone ​

ZenMux ​

Ollama Cloud ​

Local Models ​

Ollama ​

LM Studio ​

llama.cpp ​

Custom Providers ​

Custom Base URL ​

Model Selection Guide ​

Troubleshooting ​

Related Resources ​

Model Provider List

📝 Course Notes

Configuration

OpenCode Zen (Recommended for Beginners)

International Models

Anthropic Claude

OpenAI

Google Gemini

xAI Grok

Mistral

Cohere

Perplexity

GitHub Copilot

DeepSeek

Chinese Models

Moonshot AI

MiniMax

Z.AI (Zhipu)

Cloud Platforms

Amazon Bedrock

Azure OpenAI

Azure Cognitive Services

Cloudflare Workers AI

GitLab

Aggregation Platforms

OpenRouter

Groq

Cerebras

Fireworks AI

Deep Infra

Together AI

Hugging Face

Baseten

Cortecs

Nebius Token Factory

IO.NET

Venice AI

OVHcloud AI Endpoints

SAP AI Core

Cloudflare AI Gateway

Vercel AI Gateway

Helicone

ZenMux

Ollama Cloud

Local Models

Ollama

LM Studio

llama.cpp

Custom Providers

Custom Base URL

Model Selection Guide

Troubleshooting

Related Resources