Sakana AI

Fugu Mini

Q: What is Fugu Mini?

Fugu Mini is Sakana AI's AI model — Fugu's fast mini variant — single model, no panel, ~3 min per build. It has a Sakana subscription · same key as Ultra context window and was released 2026-06-15.

Q: How good is Fugu Mini at coding and one-shot builds?

On the GoldieBench one-shot build benchmark it averages 7.75/10 across 36 scored tasks, with 2 gold, 0 silver and 1 bronze medals.

Q: How much does Fugu Mini cost?

Same Sakana subscription pool as Fugu Ultra. The non-Ultra `fugu` model on Sakana's API. Sakana describes it as 'Fast mini model optimized for low latency yet high quality responses.' Crucially: zero orchestration tokens per call (vs Ultra's panel of thousands). Re

Q: Where can I see Fugu Mini demos?

Every one-shot build is live and playable on this page and on the GoldieBench compare matrix — same prompt as every other model, no retries.

Fugu's fast mini variant — single model, no panel, ~3 min per build.

ContextSingle-model variant of Sakana's Fugu — no panel orchestration. Same API endpoint, much faster per call.

PricingSame Sakana subscription pool as Fugu Ultra

Tasks tested37

Avg score7.75/10 average

Medals🥇2 🥈0 🥉1

Release2026-06-15

Official sitesakana.ai ↗

Official vendor source

Fugu Mini is built by Sakana AI — see the vendor's own product page, pricing, and docs at sakana.ai.

Visit sakana.ai →

What is Fugu Mini?

Fugu Mini is the Sakana AI frontier model with a Single-model variant of Sakana's Fugu — no panel orchestration. Same API endpoint, much faster per call. context window, released 2026-06-15. Tagline: Fugu's fast mini variant — single model, no panel, ~3 min per build.. Official source: sakana.ai.

Pricing detail. The non-Ultra `fugu` model on Sakana's API. Sakana describes it as 'Fast mini model optimized for low latency yet high quality responses.' Crucially: zero orchestration tokens per call (vs Ultra's panel of thousands). Returns in ~3 min instead of 6-15 min and doesn't time out on heavy prompts.

How I use it inside the Agent OS. Dispatched from Agent OS as the fast Sakana lane. Bench scored by Claude judge against the same 42 prompts.

What I built with Fugu Mini

Every model on Goldie Bench gets the same fixed prompt set — one shot, single HTML file out — and I score the result 0–10 inside the Agent Operating System. Here's what Fugu Mini shipped on the bench: 37 one-shot demos across Single-model variant of Sakana's Fugu — no panel orchestration. Same API endpoint, much faster per call. of context. Of those, 36 are scored against the field with my honest 0–10 from the source guides at agentos.guide.

Strengths

Zero panel orchestration — much lower latency than Ultra
Same Sakana subscription, no extra cost
Doesn't time out on heavy game/3D prompts where Ultra stalls

Trade-offs

Single model only — no ensemble verdict
Newer than Ultra — less calibration / verification

Best for

Agent loops where latency matters more than panel consensus
Quick first-drafts you'll refine downstream
Filling out a bench when Ultra is timing out

Every benchmark — Fugu Mini's full scorecard

All 36 scored tasks, best first — the judge's 0–10 on the same rubric as the whole field. Click any bar for that task's cross-model page, or open this scorecard in the interactive graphs. Full editorial breakdown with judge quotes and sourced outside research: the Fugu Mini deep dive →.

Every demo by Fugu Mini

37 live demos, sorted by category. Click any tile to play the actual one-shot result. Verdicts and 0–10 scores are pulled from the source guides where I posted them publicly.

Fugu Mini

What is Fugu Mini?

What I built with Fugu Mini

Strengths

Trade-offs

Best for

Every benchmark — Fugu Mini's full scorecard

Every demo by Fugu Mini

Compare Fugu Mini against every other model

Quick pill index

Fugu Mini — frequently asked

Run this stack yourself.