How to Use Ling 3.0 Tiny in Claude Code
Ling 3.0 Tiny suits focused coding tasks, document questions, and tool-based automation. It combines 7.9B total parameters with 1.3B active per token and supports both direct answers and multistep reasoning.
Prompts to try with Ling 3.0 Tiny in Claude Code
Copy an example into Claude Code, or ask in your own words.
Use Ling 3.0 Tiny with thinking enabled to investigate the failing test in this project. Trace the relevant function, make the smallest justified fix, and run the affected test.
Use Ling 3.0 Tiny to explain the selected function and its callers. Return its inputs, outputs, side effects, and three edge cases, with file references.
Use Ling 3.0 Tiny to draft tests for the selected parser. Cover empty input, malformed input, and one valid example using the project's existing test conventions.
Use Ling 3.0 Tiny to compare this implementation with the requirements in the project documentation. Report each mismatch with the relevant code and requirement.
Use Ling 3.0 Tiny with thinking enabled, temperature 1.0, and top_p 0.95 to review this patch for logic errors. Return only findings supported by the code and a suggested verification for each.
How to set up Ling 3.0 Tiny in Claude Code
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the hosted server
This adds Spiral to the current project. Add --scope user to use it in every project.
claude mcp add --transport http spiral https://api.spiral.ai/mcp
- Sign in
Start claude, type /mcp, choose spiral and Authenticate. Sign in to Spiral in the browser and click Authorize. No API key.
Use the Spiral skill
- Send Claude Code 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 Ling 3.0 Tiny
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": "inclusionai/ling-3.0-tiny", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Claude Code finds Ling 3.0 Tiny on Spiral, shows the price and runs it.
Ling 3.0 Tiny providers and prices
Spiral runs Ling 3.0 Tiny through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.000241 per run at the default inputs. $0.0195 per 1M input tokens, $0.0585 per 1M output tokens.
Settings you can ask for
Ask Claude Code for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | 1 to 250000 | The most output tokens to generate. On a reasoning model the reasoning tokens count toward it: a cap too low for the effort can end with no answer (finish_reason length). | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Frequency penalty | -2 to 2 |
Every API field of inclusionAI: Ling 3.0 Tiny is on its tool page: open its card above.
What can you do with Ling 3.0 Tiny in Claude Code?
Ling 3.0 Tiny can help turn a failing test or unfamiliar function into a focused coding task. Supply the relevant project files and a clear acceptance check.
Small bug fixes
Ask Ling 3.0 Tiny to trace a failing function, propose a narrow patch, and identify a test that checks the reported behavior.
Questions about internal documentation
Provide relevant documentation passages and ask for an answer with supporting excerpts and explicit gaps in the evidence.
Web copy translation
Translate page text while preserving product names, headings, and supplied terminology. Review the result against the original.
Support request triage
Assign incoming requests to a short list of categories using written rules and examples. Return ambiguous requests for review.
Meeting action extraction
Turn a transcript into actions, owners, and due dates. Require missing details to remain unknown.
Rule-based calculations
Apply supplied formulas and constraints to a calculation, returning the result with enough working to check it.
Tool-assisted information lookup
Let Ling 3.0 Tiny request a search or lookup tool, then summarize the returned evidence with source references.
More on Ling 3.0 Tiny and Claude Code
Overview
Ling 3.0 Tiny in other agents
Questions about Ling 3.0 Tiny in Claude Code
Can Claude Code use Ling 3.0 Tiny?
Yes. Add Spiral to Claude Code as an MCP server, and Claude Code can find and run Ling 3.0 Tiny like any built-in tool.
How much does Ling 3.0 Tiny cost in Claude Code?
Up to $0.000241 per run at the default inputs. $0.0195 per 1M input tokens, $0.0585 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Claude Code plan is separate.
Do I need an OpenRouter account to use Ling 3.0 Tiny in Claude Code?
No. Spiral runs Ling 3.0 Tiny on its own OpenRouter account. You sign in to Spiral once and never paste a key.
How long does it take to set up Ling 3.0 Tiny in Claude Code?
About two minutes: add the hosted server, then sign in.
Does Claude Code wait while Ling 3.0 Tiny runs?
Ling 3.0 Tiny runs in the background. Claude Code starts the run, checks on it and gives you the text, json when it's ready.
What is Ling 3.0 Tiny useful for?
Use it for focused coding assistance, questions grounded in supplied documents, and tasks that involve tools. The maker also documents mathematical reasoning and instruction following.
Does Ling 3.0 Tiny always need thinking mode?
It supports both direct responses and multistep reasoning. Try direct responses for straightforward extraction or rewriting, and thinking mode for debugging or problems with several constraints.
How much context can Ling 3.0 Tiny handle?
SGLang documents 128K native context, while Vercel advertises a 256K context window. Treat 256K as a documented extended capacity rather than the native baseline.
How long can a Ling 3.0 Tiny response be?
Vercel documents up to 32K output tokens. That is a separate limit from the context window, so a large input allowance does not imply an equally large response.
Which sampling settings should I start with?
For thinking mode, the documented starting point is temperature 1.0 and top-p 0.95. Evaluate those settings on representative tasks before changing them.
Start using Ling 3.0 Tiny in Claude Code
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Claude Code:
set up https://spiral.ai/SKILL.md


