Skip to main content
AI Choice Engine

Sakana

Sakana Fugu Ultra v2.0

Sakana Fugu Ultra v2.0 is a Sakana model with Not verified context, $5 in / $30 out per million tokens, last verified October 25, 2026. Open weights: no. Public-board scores are linked to the publisher; vendor-only figures stay in the watch-out, not on the board.

Fugu Ultra is Sakana's multi-agent API orchestration service, selecting and switching among underlying models.

Recorded profile verification October 25, 2026Individual provider fields may change

AI model specification details
SpecificationSakana Fugu Ultra v2.0
ProviderSakana
TierBalanced
Context windowNot verifiedUnverified
Max outputNot verifiedUnverified
Input / 1M tokens$5
Output / 1M tokens$30
WeightsClosed
ParametersNot verifiedUnverified
Reasoning levelshigh, xhigh, max
Modalitiestext, image
API model idfugu-ultra-v2.0
ReleasedSeptember 11, 2026

Pricing tiers: fugu-ultra-v2.0 fixed pay-as-you-go $5/$30 per MTok at context ≤272K and $10/$45 above; cached input $0.50/$1. Orchestration input/output/cache tokens are extra real usage billed at the same respective rates. Final-response token limits do not cap orchestration work. API-platform subscriptions are $20/$100/$200 monthly with relative quotas, separate from Sakana Chat limits.

Best for

  • Model routing
  • Orchestration
  • Cost optimisation

Watch out

Pinned Ultra v2.0 is a multi-model orchestration service. high/xhigh are distinct levels; max aliases xhigh here, unlike Ultra v1.1 where max is distinct. Responses defaults to xhigh for Ultra; Chat Completions documents high. previous_response_id is rejected: send full history. temperature is accepted but ignored, and output limits apply only to the final model. EU/EEA access is unavailable. Provider-shipped client metadata advertises 1M context and image input; an exact server-enforced combined-context or output maximum remains unverified.

Source receipts

Catalog figures for this model were checked against the following sources.

Common questions

Sakana Fugu Ultra v2.0

Answered from the verified figures on this page rather than general guidance.

How much does Sakana Fugu Ultra v2.0 cost per million tokens?
Sakana Fugu Ultra v2.0 lists base rates of $5 per million input tokens and $30 per million output tokens. Estimate both input and output token volume and any additional charges for your workload. fugu-ultra-v2.0 fixed pay-as-you-go $5/$30 per MTok at context ≤272K and $10/$45 above; cached input $0.50/$1. Orchestration input/output/cache tokens are extra real usage billed at the same respective rates. Final-response token limits do not cap orchestration work. API-platform subscriptions are $20/$100/$200 monthly with relative quotas, separate from Sakana Chat limits.
What is Sakana Fugu Ultra v2.0 best for?
Sakana Fugu Ultra v2.0 is a balanced tier from Sakana. It suits model routing, orchestration, cost optimisation. Pinned Ultra v2.0 is a multi-model orchestration service. high/xhigh are distinct levels; max aliases xhigh here, unlike Ultra v1.1 where max is distinct. Responses defaults to xhigh for Ultra; Chat Completions documents high. previous_response_id is rejected: send full history. temperature is accepted but ignored, and output limits apply only to the final model. EU/EEA access is unavailable. Provider-shipped client metadata advertises 1M context and image input; an exact server-enforced combined-context or output maximum remains unverified.