How to Use Qwen3.5-35B-A3B in Codex
Use Qwen3.5-35B-A3B for code changes, multilingual questions, and reasoning over long documents. Qwen reports 69.2 on SWE-bench Verified, compared with 72.4 for Qwen3.5-27B, so test proposed fixes before accepting them.
Prompts to try with Qwen3.5-35B-A3B in Codex
Copy an example into Codex, or ask in your own words.
Use Qwen3.5-35B-A3B to investigate the failing tests in this project. Trace the failure to the relevant code, make the smallest justified fix, add a regression test, and run the affected tests.
Use Qwen3.5-35B-A3B to review the current diff. Focus on behavior changes, boundary conditions, and missing tests. Give each finding a file reference and a concrete reproduction.
Use Qwen3.5-35B-A3B to trace how this project's authentication middleware handles an expired session. Cite the relevant files, identify untested branches, and propose tests before editing code.
Use Qwen3.5-35B-A3B to implement keyboard navigation for the existing dropdown component. Follow the project's conventions, use temperature 0.6 for the coding request, and verify focus movement and selection.
Use Qwen3.5-35B-A3B to compare the API documentation with the route handlers in this repository. List mismatched request requirements and response shapes with file references, then propose focused corrections.
How to set up Qwen3.5-35B-A3B in Codex
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the hosted server
codex mcp add spiral --url https://api.spiral.ai/mcp
- Sign in
Codex opens the browser: sign in to Spiral and click Authorize. No API key. codex mcp list then shows Spiral.
codex mcp login spiral
Use the Spiral skill
- Send Codex this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run Qwen3.5-35B-A3B
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openrouter", "model": "qwen/qwen3.5-35b-a3b", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Codex finds Qwen3.5-35B-A3B on Spiral, shows the price and runs it.
Qwen3.5-35B-A3B providers and prices
Spiral runs Qwen3.5-35B-A3B through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.005332 per run at the default inputs. $0.182 per 1M input tokens, $1.30 per 1M output tokens.
Settings you can ask for
Ask Codex for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Min p | 0 to 1 | ||
| Top k | a whole number from 0 | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | 1 to 81920 | The most output tokens to generate. | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Frequency penalty | -2 to 2 | ||
| Repetition penalty | 0 to 2 |
Every API field of Qwen: Qwen3.5-35B-A3B is on its tool page: open its card above.
What can you do with Qwen3.5-35B-A3B in Codex?
Turn a failing test or unfamiliar code path into a proposed fix with Qwen3.5-35B-A3B. Ask the agent to inspect the relevant project files, make a focused change, and verify the result with tests.
Fix a reproducible bug
Give Qwen3.5-35B-A3B the failing test, error output, and relevant code. Ask for a focused patch and a regression test, then run the tests.
Review a proposed code change
Ask Qwen3.5-35B-A3B to inspect a diff for behavior changes, missed edge cases, and gaps in test coverage. Require each finding to identify the affected code.
Compare technical requirements
Have Qwen3.5-35B-A3B compare supplied specifications and identify conflicting requirements. Request section references and a separate list of unresolved questions.
Translate support responses
Use Qwen3.5-35B-A3B to draft responses in the customer's language. Supply the approved answer and a glossary so the task stays tied to your support policy.
Work through a technical problem
Ask Qwen3.5-35B-A3B to solve a logic or programming problem with explicit assumptions, a concise explanation, and a checkable final result.
Prepare a recurring change brief
Have Qwen3.5-35B-A3B compare newly supplied updates with the previous brief. Ask it to identify changed requirements, affected teams, and evidence for each change.
Turn a procedure into a task plan
Give Qwen3.5-35B-A3B a written procedure and ask for ordered actions, dependencies, and completion checks. Review the plan against the original procedure.
More on Qwen3.5-35B-A3B and Codex
Overview
Qwen3.5-35B-A3B in other agents
Questions about Qwen3.5-35B-A3B in Codex
Can Codex use Qwen3.5-35B-A3B?
Yes. Add Spiral to Codex as an MCP server, and Codex can find and run Qwen3.5-35B-A3B like any built-in tool.
How much does Qwen3.5-35B-A3B cost in Codex?
Up to $0.005332 per run at the default inputs. $0.182 per 1M input tokens, $1.30 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Codex plan is separate.
Do I need an OpenRouter account to use Qwen3.5-35B-A3B in Codex?
No. Spiral runs Qwen3.5-35B-A3B on its own OpenRouter account. You sign in to Spiral once and never paste a key.
How long does it take to set up Qwen3.5-35B-A3B in Codex?
About two minutes: add the hosted server, then sign in.
Does Codex wait while Qwen3.5-35B-A3B runs?
Qwen3.5-35B-A3B runs in the background. Codex starts the run, checks on it and gives you the text, json when it's ready.
What is Qwen3.5-35B-A3B useful for?
Its documented task areas include coding, reasoning, and agent work. Practical starting points include debugging a failing test, comparing requirements across documents, and answering multilingual questions.
How much text can Qwen3.5-35B-A3B handle?
The native context window is 262,144 tokens, including input and output. Keep the supplied material below that total so the model has room to answer.
Which languages does Qwen3.5-35B-A3B support?
Qwen reports support for 201 languages and dialects. For translation work, specify the target language, audience, and terms that must remain unchanged.
Which settings should I start with for coding?
For precise coding in thinking mode, Qwen recommends temperature 0.6, top-p 0.95, top-k 20, and minimum probability 0. Start there when evaluating code changes against your tests.
What should I double-check in its coding answers?
Run the relevant tests and inspect the proposed changes. Qwen reports 69.2 on SWE-bench Verified, which is evidence of useful coding ability with remaining failures.
Start using Qwen3.5-35B-A3B in Codex
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Codex:
set up https://spiral.ai/SKILL.md


