# Long-context Question Answering on Longbench Chat

Source: [Long-context Question Answering on Longbench Chat](<https://www.sota2.com/research/sota/long-context-question-answering-on-longbench-chat>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:04:07.695Z

Evaluation of model performance on the Longbench Chat dataset for Long-context Question Answering, measured using metrics C, CLQA, and CR.

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 Question Answering |
| Dataset (including any stated split) | Longbench Chat |
| Metric | Score C |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [Score C](<https://www.sota2.com/research/sota/long-context-question-answering-on-longbench-chat.md?metric=Score%20C>): Higher is better; unit: PERCENT.

- [CLQA Score](<https://www.sota2.com/research/sota/long-context-question-answering-on-longbench-chat.md?metric=CLQA%20Score>): Higher is better; unit: PERCENT.

- [CR Score](<https://www.sota2.com/research/sota/long-context-question-answering-on-longbench-chat.md?metric=CR%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Score C | Result date | Report paper |
| --- | --- | --- | --- |
| Claude-3-sonnet | 73.8 | Not recorded | 68fa74bf1e209131cea637a4 |
| GLM-4 | 69.4 | Not recorded | 68fa74bf1e209131cea637a4 |
| LongCite-8B | 69 | Not recorded | 68fa74bf1e209131cea637a4 |
| LongCite-9B | 67.6 | Not recorded | 68fa74bf1e209131cea637a4 |
| Mistral-Large-Instruct | 63.8 | Not recorded | 68fa74bf1e209131cea637a4 |
| GPT-4o | 61.6 | Not recorded | 68fa74bf1e209131cea637a4 |
| Llama-3.1-70B-Instruct | 60.8 | Not recorded | 68fa74bf1e209131cea637a4 |
| GLM-4-9B-chat | 60.4 | Not recorded | 68fa74bf1e209131cea637a4 |
| Llama-3.1-8B-Instruct | 53.2 | Not recorded | 68fa74bf1e209131cea637a4 |

### Claude-3-sonnet

Score C: 73.8. Result ID: 699365d53a6c31484915a515.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### GLM-4

Score C: 69.4. Result ID: 699365d53a6c31484915a516.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74521e209131cea610b6.md>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### LongCite-8B

Score C: 69. Result ID: 699365d53a6c31484915a51b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

[Implementation](<https://github.com/THUDM/LongCite>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### LongCite-9B

Score C: 67.6. Result ID: 699365d53a6c31484915a51c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

[Implementation](<https://github.com/THUDM/LongCite>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### Mistral-Large-Instruct

Score C: 63.8. Result ID: 699365d53a6c31484915a51a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d61e209131cea4d909.md>)

[Implementation](<https://github.com/mistralai/mistral-src>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### GPT-4o

Score C: 61.6. Result ID: 699365d53a6c31484915a514.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### Llama-3.1-70B-Instruct

Score C: 60.8. Result ID: 699365d53a6c31484915a519.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### GLM-4-9B-chat

Score C: 60.4. Result ID: 699365d53a6c31484915a517.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74521e209131cea610b6.md>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

### Llama-3.1-8B-Instruct

Score C: 53.2. Result ID: 699365d53a6c31484915a518.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea637a4.md>)

Recorded evaluation configuration: {"Citation Strategy":"LAC-S"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.