FastMetal
Back to Models
Sakana AI
fugu-max

Fugu Max

Fugu Max is the cost-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to route...

9/11/2026
1,000,000 tokens
Input: $2.20/M
Output: $6.60/M
Try in ChatGet API Key

Specifications

Modalities

Input
text
image
file
Output
text

Supported Parameters

include_reasoning
reasoning
reasoning_effort
structured_outputs
tool_choice
tools
web_search_options

Max Output Tokens

128,000

Reasoning Configuration

Default
Thinking on (effort: high)
Selectable effort levels
max
xhigh
high
Turning thinking off
Not possible (always thinks)

How to control thinking, and what it costs

Data policy

Prompt retention
May be retained
Training
Not used for training

Whether a provider trains on prompts is a declared value from our terms with them. Retention is determined from the upstream listing for every host this model can reach. Neither is guessed.

Measured performance

Time to first response (p50)3.9s
Throughput (p50)1248 tokens/s

    Measured through FastMetal's own gateway over the last 30 days (643 first-response samples).

    Throughput is sampled once a day: the median of 27 observations over the last 30 days.

    Time to first response is measured to the first streamed chunk; throughput counts all output tokens, including reasoning tokens. These are observations, not a performance guarantee.

    Code Examples

    curl https://api.fastmetal.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "fugu-max",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

    How Fugu Max actually answers

    Real responses to our standard prompts, recorded on FastMetal.

    Count the number of 'r's in 'strawberry'

    Count the number of 'r's in 'strawberry'. Explain your reasoning step by step.

    The word **“strawberry”** has **3** letters **“r”**. Step by step: 1. Write the word: **s t r a w b e r r y** 2. Check each letter: - s — no - t — no - **r** — yes, 1st - a — no - w — no - b — no - e — no - **r** — yes, 2nd - **r** — yes, 3rd - y — no So, the total number of **‘r’s** in **“strawberry”** is **3**.
    3.7s response261 output tokens

    Debug This Error

    I'm getting the following error in my Node.js application: TypeError: Cannot read properties of undefined (reading 'map') at UserList (/app/components/UserList.js:12:25) at renderWithHooks (/app/node_modules/rea…

    The error means `users` is `undefined` when `UserList` renders, so `users.map(...)` throws: ```js function UserList({ users }) { // if `users` is not passed, this is undefined return ( <ul> {users.map(user => <li key={user.id}>{user.name}</li>)} </ul> ); } ``` Common causes: 1. Parent renders `<UserList />` without the `users` prop. 2. State is initialized as `undefined`: ```js const [users, setUsers] = useState(); ``` 3. Async data hasn’t loaded yet, but the component already renders. 4. The API returned something unexpected, e.g. `undefined`, an object, or an error response. ## Fixes ### 1. Default the prop to an empty array ```jsx function UserList({ use…
    10.6s response1293 output tokens

    Code Review

    Please review the following Python function and suggest improvements for readability, performance, and best practices: def get_data(url, retries=3): import requests import time for i in range(retries):…

    Here’s a review with concrete improvements. ## Main issues 1. **Bare `except:`** - Catches everything, including `KeyboardInterrupt`, `SystemExit`, and programming errors. - Catch `requests.RequestException` instead, plus `ValueError` for JSON parsing if needed. 2. **Retries everything, including non-retryable errors** - A `404` or `401` will be retried even though retrying will not help. - Retry only transient failures: network errors, `429`, and `5xx`. 3. **Fixed 2-second sleep** - No exponential backoff. Repeated retries can hammer the server. - Use exponential backoff, optionally with jitter. 4. **Silent failure** - Returning `[]` after all retries makes it hard…
    24.7s response3795 output tokens

    Compare these answers side by side with other models →

    Frequently asked questions

    How much does the Fugu Max API cost?
    On FastMetal, Fugu Max is billed per token: $2.20 per 1M input tokens and $6.60 per 1M output tokens on a US-dollar account, before tax. Usage is drawn from a prepaid balance; there is no subscription or monthly fee.
    Can I call Fugu Max with the OpenAI SDK?
    Yes. Point base_url at https://api.fastmetal.ai/v1 and pass "fugu-max" as the model; existing OpenAI-style code works unchanged, including streaming, tool calls and structured output.
    What is the context window of Fugu Max?
    1,000,000 tokens, shared between the prompt and the response.
    What do I need to try Fugu Max?
    Create an account and add credit; the model is then available both in the browser chat and over the API. There is no contract or minimum spend.

    Try Fugu Max right now

    Fugu Max is available on FastMetal through one API key. Start in the browser, or call it from the OpenAI SDK.

    Create a free account

    Compare with other models

    See all comparisons →

    More from Sakana AI

    Fugu Ultra v2
    1,000,000 tokens