CognitionSWE-2
As of 13 Nov 2026, SWE-2 by Cognition is not yet ranked on the BenchLeader Index, which needs at least five independent results across two categories. Released 10 Sep 2026.
Key facts (as of 13 Nov 2026)
- SWE-2 was released on 10 Sep 2026 (Cognition).
In detail. SWE-2 is a Cognition model, released 10 Sep 2026. It is not yet ranked: the index needs at least five independent results across two categories. It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 10 Sep 2026.
- Blended price
- –
- Output speed
- –
- First answer
- –
- Context
- –
- Full answer
- –
- Cost per task
- –
- Tokens per task
- –
- Released
- 10 Sep 2026
- Coding65
Versions
Cognition has shipped 3 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| maxbest | – | – | – | – | – | Coding 65 |
| not stated | – | – | – | – | – | Agents & tools 62 |
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Coding
| Benchmark | max | Source |
|---|---|---|
| FrontierCode | 50.0%#8 | Cognition |
Agents & tools
| Benchmark | not stated | Source |
|---|---|---|
| HiL-Bench | 61.5%#9 | Scale AI SEAL |
See also
Questions people ask
- Why does SWE-2 have no BenchLeader Index score?
- The index needs at least five independent results across at least two categories, and as of 13 Nov 2026 SWE-2 does not have them yet. Launch-post scores are never used, so it is ranked once independent publishers have measured it.
- When was SWE-2 released?
- Cognition released SWE-2 on 10 Sep 2026. BenchLeader has results for it at 2 reasoning-effort settings.
Data as of 13 Nov 2026. Compare these configurations.
Cite as: BenchLeader, “SWE-2: benchmarks, pricing, speed and rank”, data as of 13 Nov 2026, https://www.benchleader.com/models/swe-2.