Open-Weight Models Close the Gap With Closed Frontier Labs
Over the next 3-9 months, open-weight releases with downloadable weights, long context, and strong agentic/coding performance increasingly match closed frontier models on practical work, eroding the closed-lab moat.
weakening · confidence 67 · +1 7d · Medium term (3-9 months) · tracking since June 17, 2026 · updated July 7, 2026
Score history
Daily conviction score, 0 to 100. Higher means the thesis is more strongly corroborated.
Now 67 · -2 since Jul 6 · ranged 67 to 69
Showing the last few days. Unlock full score history.
Why the conviction moved
- Jul 4Strengthened +4
Mistral's Apache-licensed Leanstral 1.5 is a downloadable open-weight model that saturates a formal-math benchmark, extending open releases into specialized frontier-grade reasoning.
- Jul 2Strengthened +5
Moonshot's open-weight Kimi K2.7 Code being promoted to GA inside GitHub Copilot shows a downloadable model treated as a first-class coding option alongside closed frontier models.
Showing the last 2 days. Unlock the full record.
Source trail
Supporting · July 4, 2026
Mistral Open-Sources Leanstral 1.5, Saturating a Formal-Math Benchmark
Mistral's Apache-licensed Leanstral 1.5 is a downloadable open-weight model that saturates a formal-math benchmark, extending open releases into specialized frontier-grade reasoning.
AI ML Big Data (Telegram)Supporting · July 2, 2026
Moonshot's Kimi K2.7 Code Reaches General Availability in GitHub Copilot
Moonshot's open-weight Kimi K2.7 Code being promoted to GA inside GitHub Copilot shows a downloadable model treated as a first-class coding option alongside closed frontier models.
GitHub Changelog
Unlock full source trail, score history, and daily updates.
15 more sources in the full trail.
Unlock Trends