$ diff fugu-max gpt-image-2-5-flare
Fugu Max vs GPT-Image-2.5 Flare
Values come from Fugu Max and GPT-Image-2.5 Flare, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing.
Sakana AIFugu Max
- context
- 1M
- weights
- —
- $/M in
- $2
- $/M out
- $6
- context
- —
- weights
- closed
- $/M in
- $30
- $/M out
- $2
What actually differs
- Input price
- Fugu Max at $2/M against $30/M — 15.0× cheaper to feed.
- Output price
- GPT-Image-2.5 Flare at $2/M against $6/M — 3.0× cheaper to generate. Output dominates the bill on most agentic workloads, where the model writes far more than it reads.
- Recency
- Fugu Max shipped 3 days after GPT-Image-2.5 Flare (2026-09-11 vs 2026-09-08).
Full spec
| Attribute | Fugu Max | GPT-Image-2.5 Flare |
|---|---|---|
| Developer | Sakana AI | OpenAI |
| Released | 2026-09-11 | 2026-09-08 |
| Context window | 1,000,000 | not recorded |
| Pricing | $2/M input · $6/M output | image input $8.00/M tokens ($2.00/M cached) · image output $30.00/M tokens · text input $5.00/M tokens ($1.25/M cached) |
| License | not recorded | proprietary |
| Availability | Sakana API (OpenAI-compatible), OpenRouter, NanoGPT, Kilo Gateway, LLM Gateway | OpenAI API |