AI Coding Agents · OpenCode v2+ · Claude Code · OpenClaw

OpenCode Free Models, Usage Limits & opencode.json Config Generator

Wondering "is OpenCode free?" or stuck on the Free usage limit reached lockout in OpenCode Zen? Here is how OpenCode's built-in free tier actually works in 2026—plus an interactive $0 BYOK Config Generator to connect external card-free LLM APIs in 10 seconds.

1. How OpenCode Zen Built-in Free Models Work (and Why They Lock Out)

Select Free BYOK Provider for OpenCode Default Free Coding Model ID Verified Free Rate Limit (No Card) Best Coding Agent Use Case
gemini-2.5-flash
10 RPM · 500 RPD · 250K TPM (No Credit Card)
Best all-around free Coding Agent fallback with 1M context window
openai/gpt-oss-120b
30 RPM · 1,000 RPD · 200K TPD (No Credit Card)
Ultra-fast token generation for quick edits & shell commands
deepseek/deepseek-r1:free
20 RPM · 50 RPD ($0 balance) or 1,000 RPD ($10 lifetime balance)
Access to DeepSeek R1 & Qwen3 235B `:free` endpoints
codestral-latest
1 RPS (~60 RPM) · Experiment Plan (Phone verified, No Card)
Dedicated code generation model (`codestral-latest`) at $0
Meta-Llama-3.3-70B-Instruct
20 RPM · 20 RPD · 200K TPD (No Card)
Zero-data-retention backup for complex reasoning prompts

2. Generated opencode.json (One-Click Copy)

Save this in your project root or ~/.config/opencode/opencode.json:

{}

3. Shell Environment Variables (.zshrc / .bashrc)

Grab your free keys from our Free LLM API table and export them locally: