How to Use Fugu Ultra v2 in Claude Code
Use Fugu Ultra v2 for research, complex reasoning, and software development that needs several steps. It puts more emphasis on answer quality and depth than Fugu, which balances quality with responsiveness.
Prompts to try with Fugu Ultra v2 in Claude Code
Copy an example into Claude Code, or ask in your own words.
Use Fugu Ultra v2 to trace the failing checkout test through this repository. Identify the root cause, make a focused fix, and run the relevant tests.
Use Fugu Ultra v2 to review the current diff for broken contracts between the frontend and backend. Report only actionable findings with file locations and a reproducible failure scenario.
Use Fugu Ultra v2 to plan this project's database migration. Inspect the schema and callers, then list the implementation order, compatibility checks, and rollback conditions before editing.
Use Fugu Ultra v2 with xhigh reasoning to investigate the intermittent duplicate-job failure. Inspect the worker and retry code, compare plausible causes, and add a regression check for the cause you verify.
Use Fugu Ultra v2 to implement the feature described in this project's issue document. Check the existing architecture, preserve its conventions, and verify each acceptance criterion.
How to set up Fugu Ultra v2 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 Fugu Ultra v2
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": "sakana/fugu-ultra-v2", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Claude Code finds Fugu Ultra v2 on Spiral, shows the price and runs it.
Fugu Ultra v2 providers and prices
Spiral runs Fugu Ultra v2 through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.159972 per run at the default inputs. $6.50 per 1M input tokens, $39.00 per 1M output tokens.
Settings you can ask for
Ask Claude Code for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Reasoning effort | high, xhigh, or max | xhigh | How much the model reasons. |
Every API field of Sakana: Fugu Ultra v2 is on its tool page: open its card above.
What can you do with Fugu Ultra v2 in Claude Code?
Use Fugu Ultra v2 to investigate failures across a codebase and implement changes with explicit acceptance criteria. Give it the relevant project files, constraints, and a concrete way to verify the result.
Trace a bug across application layers
Follow a failure through the interface, server, and database code. Request a reproducible case, a focused patch, and checks that demonstrate the fix.
Review a substantial code change
Inspect a change for incorrect assumptions, broken contracts, and missing edge cases. Ask for findings tied to specific code and a concrete failure scenario.
Compare technical research
Compare supplied papers on their methods, evidence, and limitations. Build a claim-by-claim table and identify disagreements that need further investigation.
Explain charts in a report
Analyze supplied chart images or PDF pages. Request the relevant values, units, and trends, with unreadable labels marked as uncertain.
Plan a software migration
Turn architecture constraints and dependency information into an ordered migration plan. Include compatibility checks, rollback conditions, and acceptance criteria.
Reconcile conflicting reports
Compare supplied reports for inconsistent dates, totals, definitions, and conclusions. Produce an evidence table showing which discrepancies remain unresolved.
Produce structured evidence records
Extract claims, supporting passages, and source references from supplied material into JSON. Define how missing evidence and conflicting values should be represented.
More on Fugu Ultra v2 and Claude Code
Overview
Fugu Ultra v2 in other agents
More Claude Code guides
Questions about Fugu Ultra v2 in Claude Code
Can Claude Code use Fugu Ultra v2?
Yes. Add Spiral to Claude Code as an MCP server, and Claude Code can find and run Fugu Ultra v2 like any built-in tool.
How much does Fugu Ultra v2 cost in Claude Code?
Up to $0.159972 per run at the default inputs. $6.50 per 1M input tokens, $39.00 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 Fugu Ultra v2 in Claude Code?
No. Spiral runs Fugu Ultra v2 on its own OpenRouter account. You sign in to Spiral once and never paste a key.
How long does it take to set up Fugu Ultra v2 in Claude Code?
About two minutes: add the hosted server, then sign in.
Does Claude Code wait while Fugu Ultra v2 runs?
Fugu Ultra v2 runs in the background. Claude Code starts the run, checks on it and gives you the text, json when it's ready.
What is Fugu Ultra v2 best suited for?
Choose it for research synthesis, software development, and problems that require several connected reasoning steps. Sakana also highlights reasoning over visual and structured data.
How much material can Fugu Ultra v2 handle?
OpenRouter lists a 1,000,000-token context window and up to 128,000 completion tokens. Treat these as separate limits when planning a long analysis and its final report.
Can Fugu Ultra v2 interpret charts and PDFs?
Yes. It accepts images and PDFs and returns text, so useful tasks include explaining a chart, comparing document claims, and writing an evidence table.
Does maximum reasoning go beyond extra-high reasoning?
Sakana documents that maximum reasoning maps to extra-high reasoning for Fugu Ultra v2.0. Treat them as equivalent rather than expecting another increase in reasoning depth.
How should I prompt Fugu Ultra v2 for a difficult task?
State the decision or deliverable first, then provide the evidence, constraints, and acceptance criteria. Ask it to identify conflicting evidence, separate assumptions from findings, and check the result against those criteria.
Start using Fugu Ultra v2 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


