BlogLLM Benchmark Scores as Published, Side by Side
FinSearchComp: Results as Published
Every result the catalog holds for FinSearchComp: 9 models from 5 developers, read from 2 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Web Search and ResearchMaintainer ByteDance Seed and Columbia Business SchoolIntroduced 2025Data version 2026103001, as of 2026-10-30
FinSearchComp tests whether a search agent can finish realistic financial lookup tasks. It has 635 questions, written with 70 financial experts, covering global markets and Greater China. The three families are time-sensitive data fetching (T1), simple historical lookup (T2), and complex historical investigation (T3). Many items have an expert-marked acceptable range rather than one exact string, and a language model judges the reply against that rubric, allowing small differences such as rounding.
FinSearchComp T3
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-5OpenAI, released 2025-08-07 | 48.5% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Effort: high; Tools: enabled; search, code interpreter, web browsing; Context: 256000 |
| Claude Sonnet 4.5Anthropic, released 2025-09-29 | 44% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Effort: thinking; Tools: enabled; search, code interpreter, web browsing; Context: 256000 |
| DeepSeek-V3.2-ExpDeepSeek, released 2025-09-29 | 27% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Tools: enabled; search, code interpreter, web browsing; Context: 256000 |
| Kimi K2 ThinkingMoonshot AI, released 2025-11-06 | 47.4% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Effort: thinking; Tools: enabled; search, code interpreter, web browsing; Context: 256000 |
| Kimi K2 Instruct 0905Moonshot AI, released 2025-09-05 | 10.4% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Tools: enabled; search, code interpreter, web browsing; Context: 256000 |
FinSearchComp T2 and T3
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Opus 4.5Anthropic, released 2025-11-24 | 66.2% | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15 | Tools: search, code interpreter, browsing |
| Gemini 3 ProGoogle, released 2025-11-18 | 49.9% | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15 | Effort: high; Tools: search, code interpreter, browsing |
| DeepSeek-V3.2DeepSeek, released 2025-12-01 | 59.1% | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15 | Tools: search, code interpreter, browsing |
| Kimi K2.5Moonshot AI, released 2026-01-27 | 67.8% | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15 | Tools: search, code interpreter, browsing |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.