Evolutionary model merge / research models is a Research from Sakana AI (Japan).
One-line fit: best used for Scientific novelty.
Key specifications
| Provider | Sakana AI |
|---|---|
| Model | Evolutionary model merge / research models |
| Type | Research |
| Context window | Varies |
| Pricing (indicative) | Research |
| Company category | Research Lab |
| Region | Japan |
| Official site | https://sakana.ai |
Benchmarks and public standing
Paper results
Treat scores as directional. SWE-bench, Terminal-Bench, LMSYS Arena, Artificial Analysis, and vendor cards use different harnesses and are not always comparable 1:1. Re-check the latest official model card before decisions.
What this model is good at
Scientific novelty
This aligns with Sakana AI’s broader strengths:
- Model merging research
- Efficient novel architectures
- Research culture
Limitations and watch-outs
Not ChatGPT competitor product
- Rate limits, regional availability, and data retention policies vary by plan.
- Agent harness quality (Cursor, Claude Code, Codex, custom tools) can change outcomes more than raw model Elo.
- Open-weight availability (if any) is separate from hosted API quality and safety filters.
Ideal users
- Product engineers shipping features that match: Scientific novelty
- Teams standardizing on the Sakana AI ecosystem
- Agent builders who need a Research
When to pick something else
- Cheaper volume: compare lower tiers from the same lab or open Chinese/EU alternatives.
- Maximum hard SWE thoughtfulness: compare Claude Fable-class and other coding flagships.
- Giant multimodal corpora / Workspace: compare Gemini-class models.
- Self-host / open weights: Llama, Qwen, DeepSeek, GLM, Mistral open lines.
Parent company
Sakana AI – Nature-inspired AI lab (founders from Google Brain); evolutionary model merging and research.
Related models from Sakana AI
Data is curated for AIForumSphere model directory mid-2026 directional figures.
