opencode.ai Free No-Key Models Complete Guide: 7 Truly Free + Zen Gateway for Zero-Cost Coding Agents (195K Stars)
TL;DR: 195K-star open-source coding agent opencode.ai offers 7+2 truly free models via Zen gateway (input/output/cached-read all $0) + local Ollama truly no-key. Verified via
GET /zen/v1/models(67 models) on 2026-09-03.
Updated: 2026-09-03 · Site https://opencode.ai · Docs https://opencode.ai/docs · Zen https://opencode.ai/zen · Models https://opencode.ai/zen/v1/models
Why opencode.ai?
- 195K stars / 950 contributors / 16M devs, 13K+ commits, TUI/Desktop/IDE/Web with multi-session
- Zen curation: tested with model teams + benchmarked providers — the right variant, not OpenRouter lottery
- Free to start: 7 Free on Zen + local $0, 75+ providers, Copilot/ChatGPT login reuse
Two kinds of "free" + 9-model list (verified 2026-09-03)
Source:
GET https://opencode.ai/zen/v1/models(67 ids) + Zen Pricing Free rows. Zen Free needs one Zen key via/connectbut bills $0; local Ollama needs no key at all.
| Model | ID | Endpoint | Highlight | Price |
|---|---|---|---|---|
| Big Pickle | big-pickle |
/zen/v1/chat/completions |
Mystery hit | Free |
| MiMo-V2.5 Free | mimo-v2.5-free |
/zen/v1/chat/completions |
Xiaomi, tool-calling | Free |
| Ling 3.0 Flash Fin Free | ling-3.0-flash-fin-free |
/zen/v1/chat/completions |
Ant, finance | Free |
| Nemotron 3 Ultra Free | nemotron-3-ultra-free |
/zen/v1/chat/completions |
NVIDIA flagship | Free |
| Nemotron 3.5 Lightning Free | nemotron-3.5-lightning-free |
/zen/v1/chat/completions |
NVIDIA lightning | Free |
| Muse Spark 1.3 Contributor Free | muse-spark-1.3-contributor-free |
/zen/v1/responses |
Meta | Free |
| Muse Spark 1.2 Contributor Free | muse-spark-1.2-contributor-free |
/zen/v1/responses |
Meta | Free |
| DeepSeek V4 Flash Free | deepseek-v4-flash-free |
/zen/v1/chat/completions |
DeepSeek flash | Free |
| Laguna S 2.1 Free | laguna-s-2.1-free |
/zen/v1/chat/completions |
New free | Free |
Full 67: OpenAI
gpt-5.6/5.5/5.4/5.3/5.2/5.1/5, Anthropicclaude-opus-5/sonnet-5/haiku-4.5/fable-5, Googlegemini-3.8~3-flash, xAIgrok-4.6, domesticdeepseek-v4/glm-5/minimax-m3/kimi-k3/qwen3.6—curl https://opencode.ai/zen/v1/models | jq .data[].id
Radar (5D)
Beyond free: paid picks (per 1M)
| Model | Input / Output / Cached | Use |
|---|---|---|
| MiniMax M3 | $0.30 / $1.20 / $0.06 | Best value after free |
| DeepSeek V4 Flash Peak | $0.44 / $1.32 / $0.014 | Peak/valley |
| Kimi K2.5 | $0.60 / $3.00 / $0.10 | Long context |
| GPT-5.1 Codex | $1.07 / $8.50 / $0.107 | Code |
| Gemini 3 Flash | $0.50 / $3.00 / $0.05 | Multimodal |
| Claude Opus 4.5 | $5.00 / $25.00 / $0.50 | Top reasoning |
Zen PAYG $20 + $1.23 fee, auto top-up $20 when < $5, monthly cap, cancel anytime.
5-min quickstart
TUI
curl -fsSL https://opencode.ai/install | bash
opencode
/connect # pick OpenCode Zen, paste key from https://opencode.ai/auth
/models # search mimo-v2.5-free
opencode.json (~/.config/opencode/): {"$schema":"https://opencode.ai/config.json","model":"opencode/mimo-v2.5-free","autoupdate":true} — swap to ollama/qwen3-coder-30b with provider.ollama.options.baseURL for offline.
API — three endpoints: /chat/completions (mimo), /responses (muse-spark), /messages (claude).
curl https://opencode.ai/zen/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer $OPENCODE_ZEN_KEY" -d '{"model":"mimo-v2.5-free","messages":[{"role":"user","content":"Write a Go crawler with rate limit"}]}'
curl https://opencode.ai/zen/v1/responses -H "Content-Type: application/json" -H "Authorization: Bearer $OPENCODE_ZEN_KEY" -d '{"model":"muse-spark-1.2-contributor-free","input":"Explain Raft in 5 bullets"}'
from openai import OpenAI
client = OpenAI(base_url="https://opencode.ai/zen/v1", api_key="OPENCODE_ZEN_KEY")
print(client.chat.completions.create(model="mimo-v2.5-free", messages=[{"role":"user","content":"retry decorator"}]).choices[0].message.content)
Picks & pitfalls
| Scenario | Free | Paid alt | | Daily | mimo-v2.5-free | MiniMax M3 | | Heavy refactor | nemotron-3-ultra-free | Claude Sonnet 4.5 | | Speed | nemotron-3.5-lightning-free | DeepSeek V4 Flash | | Finance | ling-3.0-flash-fin-free | Qwen3.5 Plus |
- Free still has rate limits (429), watch monthly cap & auto-reload
- Key at
~/.local/share/opencode/auth.json, don't commit - Zen US-hosted, zero retention, not for training — use local for privacy
Summary
Use mimo / nemotron-ultra free for daily ($0), swap to MiniMax/DeepSeek when needed, Ollama for privacy — one opencode.json switch. Workflow: free -> paid on demand -> local, <$5/mo.
Complements Kimi K2 / GLM / FLUX.1 on https://apishare.cc/free-api
🚀 Get Started: One-Click Free API Access
Want to call all the free models above with a single API key, no need to sign up for each provider? Apishare.cc provides a unified API Key — one key, 100+ models, free models at zero cost.
👉 Register on Apishare.cc → Get your unified API Key
📊 Want to see more free model rankings? Check out the Sep 2026 Free LLM API Rankings →
Keep Browsing the Rankings
- Free API rankings overview: https://apishare.cc/free-api
- Full free API directory (free quotas and rate limits): https://apishare.cc/free-api
- Create a free APIShare account: https://apishare.cc/register
- Already registered? Log in: https://apishare.cc/auth/login
- Developer documentation: https://apishare.cc/docs?utm_source=apishare_devto&utm_medium=article&utm_campaign=free_api_batch2
- More rankings: https://apishare.cc/free-api?utm_source=apishare_devto&utm_medium=article&utm_campaign=free_api_batch2
- Claim your free credit: https://apishare.cc/console?utm_source=apishare_devto&utm_medium=article&utm_campaign=free_api_batch2