Fugu Ultra
Fugu Ultra processes million-token multimodal inputs with ease.
SakanaMultimodalClosed
Function callingStructured outputsReasoningWeb searchVision
Model page Updated 2026-06-24
About Fugu Ultra
Fugu Ultra incorporates a one-million-token context window that accommodates lengthy text sequences alongside images. Sakana developed it as a proprietary system without open weights or disclosed parameter counts. This architecture supports unified handling of combined text and visual data streams.
Its primary strength lies in maintaining coherence across very long multimodal contexts. Users typically apply it to tasks requiring analysis of extended documents that include embedded images. The closed nature limits direct customization while emphasizing consistent performance on supported modalities.
Capabilities
Long-context reasoning
Multimodal text-image understanding
Vision-language integration
Extended document analysis
Image interpretation and description
How Fugu Ultra compares
Fugu Ultra (striped bar) vs other multimodal on intelligence, speed and price.
Price
USD per 1M output tokens · Lower is better · Fugu Ultra ranks #141 of 158
$25.0
$25.0
$25.0
$25.0
$30.0
$30.0
$30.0
Fugu Ultra$30.0
$30.0
$40.0
$50.0
$50.0
$50.0
Sources: Artificial Analysis (intelligence, speed) · OpenRouter (price).
Best for
Long-form Document Analysis with Visuals
Ideal for processing entire books or extensive reports containing images, using its million-token context to maintain coherence across the full content while interpreting visual elements.
Complex Vision-Language Tasks
Excels at integrating text and image understanding in scenarios requiring extended reasoning, such as analyzing infographics within lengthy articles.
Research Paper Review
Suitable for extended analysis of academic papers with figures and diagrams, enabling detailed image descriptions and logical connections over large contexts.
Strengths & limitations
Strengths
- +Supports very large 1M-token contexts
- +Native handling of both text and images
- +Suitable for complex multimodal workflows
Limitations
- –High compute demands for maximum context
- –Limited to text and image modalities
- –Less mature ecosystem compared to established models
Pricing by provider
Live per-provider pricing & uptime, routed via OpenRouter. Prices are USD per 1M tokens.
| Provider | Input /1M | Output /1M | Context | Uptime |
|---|---|---|---|---|
| Sakana AI | $5.00 | $30.00 | 1000K | 100.0% |
Cost calculator
Estimate what Fugu Ultra would cost for your usage.
$0.0200
per request
$200
estimated / month
Based on Fugu Ultra's $5.00/1M input · $30.00/1M output. Estimate only — actual cost varies by provider and caching.
Quick start
OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. Only the model slug changes between models.
JavaScript · openai
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const completion = await client.chat.completions.create({
model: "sakana/fugu-ultra",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);Model slug: sakana/fugu-ultra
Editor's verdict
Our take on Fugu Ultra
Fugu Ultra is Sakana's proprietary multimodal with a 1000K-token context window.
At $30.00 per 1M output tokens, it is premium-priced for its class, served by 1 provider.
It is available through Sakana's API and aggregators like OpenRouter.
Best suited to supports very large 1m-token contexts and native handling of both text and images.
Did you find this helpful?
Frequently asked questions
Fugu Ultra supports a context length of 1,000,000 tokens.
User reviews
Real, verified reviews from the community shape this model's rating.
Loading reviews…
Similar models
Other multimodal worth comparing.