Get the Agent OS + join 4,000+ founders inside the AI Profit Boardroom → Join AIPB ($59/mo)
Real head-to-head · same prompt, one shot

MiniMax M3 vs Fugu Ultra 1.1

1M-context frontier model at $0.30/M tokens — cheapest big-context model on the bench. vs Sakana's multi-agent orchestrator, v1.1 — routes experts per request.

MiniMax M3 · context1M tokens
Fugu Ultra 1.1 · context1M tokens
MiniMax M3 · price$0.30 / 1M input tokens, $1.50 / 1M output
Fugu Ultra 1.1 · priceAPI · orchestration billed
MiniMax M3 · vendorMiniMax
Fugu Ultra 1.1 · vendorSakana AI

What I tested — same prompt, two models

I run the same fixed prompt set through every new model the day it drops — same string, one shot, single HTML file out — and I score the result 0–10 on whether it ran, how close it hit the brief, and how good it looked. Below is what came out when I gave the exact same prompts to MiniMax M3 and Fugu Ultra 1.1, side by side, on 0 shared tasks inside the Agent Operating System.

Both models were given identical prompts inside the Agent Operating System — no help, no iteration, no "best of N" tricks. I run each prompt once, save the HTML file the model produces, and score it 0–10 on whether it ran, how close it hit the brief, and how good it looked. The scoring is mine. The verdicts below are pulled from my source comparison guides at agentos.guide where I publish every score and the reasoning behind it.

MiniMax M3 · Bench prompts dispatched via OpenRouter. Scored by Claude judge against the same 42 prompts every other model ran.

Fugu Ultra 1.1 · Benched on GoldieBench via Sakana's Responses API (fugu-ultra-v1.1, xhigh reasoning). Game tasks use the skill-infused threejs-game-director prompt plus a controls+graphics fix loop with an anti-regression clamp, judged on a real mid-play frame by the same Opus judge as the field.

Side-by-side on 47 shared tasks

Click any cell to play that model's actual one-shot attempt. Medals are derived from my 0–10 scores per task (highest = 🥇, second = 🥈, third = 🥉).

Task ↓
MiniMax M3
Fugu Ultra 1.1
Game
MiniMax M3 on Arcade
— not attempted —
Game
🥉MiniMax M3 on Crypt
— not attempted —
Game
MiniMax M3 on Dogfight
— not attempted —
Game
MiniMax M3 on Doom
— not attempted —
🥉MiniMax M3 on Dragonflight
— not attempted —
🥇MiniMax M3 on Dragonrealm
— not attempted —
Game
MiniMax M3 on Flightsim
— not attempted —
Game
MiniMax M3 on Game
— not attempted —
Game
MiniMax M3 on Gtadrive
— not attempted —
Game
MiniMax M3 on Gtafoot
— not attempted —
MiniMax M3 on Neonblaster
— not attempted —
Game
MiniMax M3 on Neoncity
— not attempted —
Game
MiniMax M3 on Neonracer
— not attempted —
🥈MiniMax M3 on Nordiccrypt
— not attempted —
Game
MiniMax M3 on Outrun
— not attempted —
Game
MiniMax M3 on Parachute
— not attempted —
Game
MiniMax M3 on Pool
— not attempted —
Game
🥇MiniMax M3 on Racing
— not attempted —
Game
MiniMax M3 on Raycaster
— not attempted —
Game
MiniMax M3 on Rpg
— not attempted —
Game
MiniMax M3 on Skyrim
— not attempted —
🥉MiniMax M3 on Twilightvale
— not attempted —
Game
MiniMax M3 on Voxelcraft
— not attempted —
Page
🥉MiniMax M3 on Aipbpromo
— not attempted —

Strengths & weaknesses I logged

MiniMax M3

Strengths

  • 1M token context — full repo / full deep-research corpus fits in one call
  • $0.30/M input is roughly 1/30th of Opus 4.8 — built for high-volume agent loops
  • Solid one-shot HTML output — clean structure on game and visual prompts

Trade-offs

  • Less polished than Fusion's panel-ensembled output on the toughest deep builds
  • Newer model — less community calibration vs Fable 5 / Opus 4.8

Fugu Ultra 1.1

Strengths

  • Orchestrates 1-3 expert agents per request and synthesises their answers
  • Reported SWE-Bench Pro 73.7 — above Opus 4.8 and GPT-5.5 on Sakana's table
  • OpenAI- and Anthropic-compatible API — drop-in for Codex and Claude Code

Trade-offs

  • Benched as a partial run until the full 50-task batch completes
  • Region-locked: unavailable across the EU/EEA/UK/CH — access depends on where you are

Pricing & context — the spec sheet

Spec MiniMax M3 Fugu Ultra 1.1
VendorMiniMaxSakana AI
Context window1,048,576-token context — matches GLM-5.2 and Fable 51,000,000-token context window
Price$0.30 / 1M input tokens, $1.50 / 1M outputAPI · orchestration billed
Pricing detailMiniMax M3 is the cheapest 1M-context frontier model on the bench — roughly 1/200th the per-call cost of OpenRouter Fusion and 1/30th of Claude Opus 4.8. Designed for high-volume agent workloads where context length matters but per-call budget is tight.Sakana's Fugu Ultra v1.1 — a multi-agent system served as a model: an orchestrator routes each request across one to three expert agents and synthesises the answer. Reported (v1.0): SWE-Bench Pro 73.7, LiveCodeBench 93.2, GPQA Diamond 95.5. NOTE: Sakana geo-blocks the EU/EEA/UK/Switzerland pending GDPR compliance — benched from an allowed region.
Release2026-06-182026-07
Bench coverage47/47 scored · avg 7.97/100/0 scored · avg —

The verdict — which should you pick?

Not enough scored shared tasks yet for a head-to-head average. The live demos for both are on the matrix above — play them and form your own opinion.

If you only run one of these inside your stack, the head-to-head average above is the call. If you can run both, my honest play is to wire MiniMax M3 and Fugu Ultra 1.1 both into the Agent Operating System and dispatch each from the kanban by task type — high-volume agent workflows where per-call cost dominates → MiniMax M3, hard, high-stakes coding and reasoning where answer quality beats latency → Fugu Ultra 1.1. That's the same setup I run for the 4,000+ founders inside the AI Profit Boardroom.

FAQ — MiniMax M3 vs Fugu Ultra 1.1

Which is better, MiniMax M3 or Fugu Ultra 1.1?

On Goldie Bench, MiniMax M3 averages no scored verdicts yet across the shared tasks, with 2 gold, 1 silver, 5 bronze overall. Fugu Ultra 1.1 averages no scored verdicts yet, with 0 gold, 0 silver, 0 bronze. Not enough scored shared tasks yet to call a winner.

How much does MiniMax M3 cost vs Fugu Ultra 1.1?

MiniMax M3: MiniMax M3 is the cheapest 1M-context frontier model on the bench — roughly 1/200th the per-call cost of OpenRouter Fusion and 1/30th of Claude Opus 4.8. Designed for high-volume agent workloads where context length matters but per-call budget is tight. Fugu Ultra 1.1: Sakana's Fugu Ultra v1.1 — a multi-agent system served as a model: an orchestrator routes each request across one to three expert agents and synthesises the answer. Reported (v1.0): SWE-Bench Pro 73.7, LiveCodeBench 93.2, GPQA Diamond 95.5. NOTE: Sakana geo-blocks the EU/EEA/UK/Switzerland pending GDPR compliance — benched from an allowed region.

What's the context window for MiniMax M3 vs Fugu Ultra 1.1?

MiniMax M3 has a 1,048,576-token context — matches GLM-5.2 and Fable 5 context window. Fugu Ultra 1.1 has a 1,000,000-token context window context window.

When should I pick MiniMax M3 over Fugu Ultra 1.1?

Pick MiniMax M3 for: High-volume agent workflows where per-call cost dominates; 1M-context tasks (whole-repo refactors, deep-research synthesis); Drop-in cheaper alternative to GLM-5.2 with comparable 1M context. The trade-off is the weaknesses we logged on the bench: Less polished than Fusion's panel-ensembled output on the toughest deep builds; Newer model — less community calibration vs Fable 5 / Opus 4.8.

When should I pick Fugu Ultra 1.1 over MiniMax M3?

Pick Fugu Ultra 1.1 for: Hard, high-stakes coding and reasoning where answer quality beats latency; Agentic workflows in Codex / Claude Code via the drop-in provider config; One-shot builds you want a panel of experts on, not a single model. The trade-off is the weaknesses we logged on the bench: Benched as a partial run until the full 50-task batch completes; Region-locked: unavailable across the EU/EEA/UK/CH — access depends on where you are.

How does Goldie Bench score MiniMax M3 vs Fugu Ultra 1.1?

Every demo on this page was built by Julian Goldie inside the Agent Operating System — same fixed prompt for both models, one shot, single HTML file out. Each result gets a 0–10 score on whether it ran, how close it hit the brief, and how good it looked. The highest score on each task gets gold; second gets silver; third gets bronze. See methodology for full provenance.

The same stack Julian uses

Run this stack yourself.

Every demo on this bench was built inside the Agent Operating System — one prompt, one shot, single HTML file out. The Agent OS, the prompts, the templates, the weekly walkthroughs and 4,000+ founders shipping with it every day all live inside the AI Profit Boardroom.

4,000+founders
258documented wins
38countries
$59/momonthly