# Long-context reasoning on BrowseComp+ 1K documents Source: [Long-context reasoning on BrowseComp+ 1K documents]() Collection index: [llms.txt]() Data updated: 2026-03-18T06:41:19.441Z Evaluation of model performance on long-context reasoning tasks using the specific 1,000 document subset of the BrowseComp+ dataset, measured primarily by accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Long-context reasoning | | Dataset (including any stated split) | BrowseComp+ 1K documents | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | SRLM (no sub-calls) | 94.6 | Not recorded | 69ba316b0636e5bf4afc05f4 | | SRLM | 92.4 | Not recorded | 69ba316b0636e5bf4afc05f4 | | RLM (no sub-calls) | 89.7 | Not recorded | 69ba316b0636e5bf4afc05f4 | | RLM | 86 | Not recorded | 69ba316b0636e5bf4afc05f4 | | Summary agent | 70.5 | Not recorded | 69ba316b0636e5bf4afc05f4 | | SRLM | 59.7 | Not recorded | 69ba316b0636e5bf4afc05f4 | | CodeAct (+ BM25) | 51 | Not recorded | 69ba316b0636e5bf4afc05f4 | | SRLM (no sub-calls) | 50.1 | Not recorded | 69ba316b0636e5bf4afc05f4 | | Summary agent | 38 | Not recorded | 69ba316b0636e5bf4afc05f4 | | RLM | 37.1 | Not recorded | 69ba316b0636e5bf4afc05f4 | | RLM (no sub-calls) | 36.3 | Not recorded | 69ba316b0636e5bf4afc05f4 | | CodeAct (+ BM25) | 12.7 | Not recorded | 69ba316b0636e5bf4afc05f4 | | Base Model | 0 | Not recorded | 69ba316b0636e5bf4afc05f4 | | CodeAct (+ sub-calls) | 0 | Not recorded | 69ba316b0636e5bf4afc05f4 | | Base Model | 0 | Not recorded | 69ba316b0636e5bf4afc05f4 | | CodeAct (+ sub-calls) | 0 | Not recorded | 69ba316b0636e5bf4afc05f4 | ### SRLM (no sub-calls) Accuracy: 94.6. Result ID: 69ba484447f34ce924a37554. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. ### SRLM Accuracy: 92.4. Result ID: 69ba484447f34ce924a37553. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. ### RLM (no sub-calls) Accuracy: 89.7. Result ID: 69ba484347f34ce924a37552. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. ### RLM Accuracy: 86. Result ID: 69ba484347f34ce924a37551. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. ### Summary agent Accuracy: 70.5. Result ID: 69ba484347f34ce924a37550. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. ### SRLM Accuracy: 59.7. Result ID: 69ba484347f34ce924a3754b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### CodeAct (+ BM25) Accuracy: 51. Result ID: 69ba484347f34ce924a3754e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. ### SRLM (no sub-calls) Accuracy: 50.1. Result ID: 69ba484347f34ce924a3754c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### Summary agent Accuracy: 38. Result ID: 69ba484347f34ce924a37548. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### RLM Accuracy: 37.1. Result ID: 69ba484347f34ce924a37549. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### RLM (no sub-calls) Accuracy: 36.3. Result ID: 69ba484347f34ce924a3754a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### CodeAct (+ BM25) Accuracy: 12.7. Result ID: 69ba484347f34ce924a37546. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### Base Model Accuracy: 0. Result ID: 69ba484347f34ce924a37545. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### CodeAct (+ sub-calls) Accuracy: 0. Result ID: 69ba484347f34ce924a37547. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-Coder-480B"}. This record does not assert independent reproduction. ### Base Model Accuracy: 0. Result ID: 69ba484347f34ce924a3754d. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. ### CodeAct (+ sub-calls) Accuracy: 0. Result ID: 69ba484347f34ce924a3754f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.