Is Beam 1.0 state of the art?#31Best recorded rank: #31 in Abstraction (ARC), as of Nov 22, 2026.
As of Nov 22, 2026, Beam 1.0 is ranked #31 in Abstraction (ARC) (since Nov 22, 2026; sources ARC-AGI; methodology v1.0). Last listed at #31 in Abstraction (ARC) on 2026-10-05. That board does not award a crown.
Facts and scores
Beam 1.0 is a closed-weights text model from Reflection AI, listed on ARC-AGI since Oct 5, 2026; no public parameter count and no public price.
How it scores
- On ARC-AGI it scores 74.0%, #54 of 123 (top 44%); 24.5 pts behind Claude Fable 5; Gemini 3 Pro Preview is 1.0 pts above and GPT-5.1 is 1.2 pts below.
- On Terminal-Bench 4.0 it scores 6.4 (CI 4.7–8.0), #39 of 44 (top 89%); 58.4 behind Claude Opus 5.5; DeepSeek V4 Pro 0813 is 0.3 above and Gemini 3.1 Pro is 3.7 below.
History
First result on isitsota: Oct 5, 2026. Best overall rank so far: #31 in Abstraction (ARC), first held Nov 22, 2026.
The ruling, in plain English
Ranked #41 in Coding.
Evidence
| Source | Score | Date | Type | Receipt |
|---|---|---|---|---|
| Terminal-Bench 4.0 Beam 1.0 | 6.4 4.7–8.0 | 2026-11-18 | INDEP. | snap_1c5180d602c4 |
| ARC-AGI beam-1.0_xhigh · xhigh | 74.0% | 2026-10-05 | INDEP. | snap_2eda3fa96d37 |
Where it sits in the succession
Crown intervals for this model are still empty.
Where it is not the crown
| Category | This model | Rank | Leader |
|---|---|---|---|
| Coding | #41 in this category | #41 | Claude Sonnet 5.5 |
| Coding › Agentic | #33 in this category | #33 | Claude Opus 5.5 |
| Abstraction (ARC) | #31 in this category | #31 | GPT-6.2 Astra |
Related pages
Is Beam 1.0 SOTA?
Best recorded rank: #31 in Abstraction (ARC), as of Nov 22, 2026.
What is Beam 1.0 best at?
As of Nov 22, 2026, Beam 1.0 is ranked #31 in Abstraction (ARC) (since Nov 22, 2026; sources ARC-AGI).
How does Beam 1.0 rank in Coding?
Best current rank: #41 in Coding, as of Nov 22, 2026.