Free Qwen3 API.
No key. No subscription.
Qwen3-Next 80B Instruct (3B active MoE, 262K context, fast) and Qwen3.5 122B (10B active). Reasoning + coding. No key.
Try it now.
No API key. No wallet. No signup. Paste this into any terminal — the response streams back from Qwen hosted free on NVIDIA, routed through BlockRun.
curl
curl https://blockrun.ai/api/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "nvidia/qwen3-next-80b-a3b-instruct",
"messages": [{"role": "user", "content": "Refactor this loop into a list comprehension: result = []\nfor x in items:\n if x > 0: result.append(x*2)"}]
}'Qwen3-Next 80B Instruct (Free)
nvidia/qwen3-next-80b-a3b-instruct
- Context
- 262K
- Price
- free
- Best for
- reasoning · coding
Qwen3.5 122B (Free)
nvidia/qwen3.5-122b-a10b
- Context
- 131K
- Price
- free
- Best for
- reasoning · coding
6 ways to use Qwen free.
BlockRun is the access layer. Pick the surface that matches how you build — terminal, notebook, IDE, agent runtime — and the same free models work everywhere.
- shell
# Install once npm install -g @blockrun/clawrouter # Then point any OpenAI-compatible client at the local proxy. # ClawRouter routes to nvidia/qwen3-next-80b-a3b-instruct (or the cheapest capable model) # without changing your code. - shell
# Add the BlockRun MCP server (Claude Code, Cursor, or ChatGPT desktop) claude mcp add blockrun -s user -- npx -y @blockrun/mcp@latest # Then call from inside the editor: # blockrun_chat(model="nvidia/qwen3-next-80b-a3b-instruct", messages=[{role:"user", content:"…"}]) - shell
curl https://blockrun.ai/api/v1/chat/completions \ -H "Content-Type: application/json" \ -d '{ "model": "nvidia/qwen3-next-80b-a3b-instruct", "messages": [{"role": "user", "content": "Refactor this loop into a list comprehension"}] }' - python
# Works with the OpenAI SDK — no key required for free models from openai import OpenAI client = OpenAI( base_url="https://blockrun.ai/api/v1", api_key="not-needed-for-free-models", ) response = client.chat.completions.create( model="nvidia/qwen3-next-80b-a3b-instruct", messages=[{"role": "user", "content": "Refactor this loop into a list comprehension"}], ) print(response.choices[0].message.content) - typescript
// Works with the OpenAI SDK — no key required for free models import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://blockrun.ai/api/v1", apiKey: "not-needed-for-free-models", }); const r = await client.chat.completions.create({ model: "nvidia/qwen3.5-122b-a10b", messages: [{ role: "user", content: "Refactor this loop into a list comprehension" }], }); console.log(r.choices[0].message.content); - 06
Franklin Agent
the AI agent with a wallet — free OSS models for routine tasks, paid models on demand
Learn more →shell# Install Franklin Agent curl -fsSL https://franklin.run/install | sh # Run with this model franklin chat --model nvidia/qwen3-next-80b-a3b-instruct "Summarize the README"
We don't share
your data.
Your prompt goes to the AI provider you picked. Nothing else, nowhere else. No training, no retention beyond the request, no profile linking.
- No training, no retention beyond the request. Your prompt is forwarded only to the AI provider you select.
- Wallet in, prompt out. Pseudonymous by default — no email, no phone number, no identity documents.
- Read the code, audit the wire format, run it yourself. @blockrun/llm and blockrun-llm on npm and PyPI.
Want Claude, GPT-5,
or Gemini too?
No subscription. No monthly minimum. Pay per call in USDC via x402 — works the same endpoint, same SDK, same model IDs. Connect a wallet, top up $5, call any frontier model. No credit card.
Everything you might
be wondering.
- Why two Qwen3 models?
- Qwen3-Next 80B Instruct is the fast workhorse (3B active params, 262K context) — great for general reasoning and coding. Qwen3.5 122B (10B active) is a step up in capability for harder tasks. Both are free, no key.
- Is Qwen3 good enough for Cursor / Continue / Claude Code?
- Yes — Qwen3-Next handles most refactoring, explain, and review tasks well, and it's free. Drop it in via ClawRouter or the BlockRun MCP server.
- How does it compare to GPT-4 / Claude on code?
- Frontier closed models are still ahead on the hardest novel problems, but Qwen3 is dramatically cheaper (free, here) and matches them on most production refactor / explain / lint workflows.
- Will my prompts be used to train Qwen?
- NVIDIA's free tier (which hosts these models) reserves the right to use prompts for service improvement. Don't send proprietary code you wouldn't paste into a public Discord. For private inference, switch to paid models.
- What happened to Qwen3 Coder 480B / Qwen3-Next Thinking?
- Both were retired from NVIDIA's free upstream. Calls to them auto-reroute to a healthy free Qwen / coding model via BlockRun's health gate, so requests still return 200.