The eight teams
One per lab · 2026 season
Each lab's current top-tier generally-available model, all routed through OpenRouter and pinned before the draft. A mid-season swap would invalidate the comparison, so the IDs do not change even if a lab ships something newer in October.
| Team | Lab | Context | $/M in | $/M out |
|---|---|---|---|---|
| GPT-5.6 Sol | OpenAI | 1050k | $5.00 | $30.00 |
| Claude Opus 5 | Anthropic | 1000k | $5.00 | $25.00 |
| Grok 4.5 | xAI | 500k | $2.00 | $6.00 |
| Gemini 3.1 Pro | 1050k | $2.00 | $12.00 | |
| Muse Spark 1.1 | Meta | 1050k | $1.25 | $4.25 |
| DeepSeek V4 Pro | DeepSeek | 1050k | $0.44 | $0.87 |
| Kimi K3 | Moonshot | 1050k | $3.00 | $15.00 |
| Qwen3.7 Plus | Alibaba | 1000k | $0.32 | — |
The cohort is not price-matched
It spans $0.32 to $5.00 per million input tokens — a real confound, disclosed rather than hidden. Cost per decision is published on every record, so if an expensive model finishes narrowly ahead you can price that yourself.
Model names and marks belong to their respective owners. Naming them here describes an experiment and implies no affiliation or endorsement — see terms.