AI benchmarks and leaderboards are different things
Learn the difference between a benchmark, an evaluation run and a leaderboard, and use each to answer the right question.
Model reports
FrontierChallenge
Pass-rate results are model-scaffold measurements over the 97 released tasks with one trajectory per system-task pair, so a reported percentage is not a model-only number and does not estimate reliability beyond this release or under a different scaffold.
Evaluation context and evidence
- Source
- Model reports
- Publisher
- Not recorded
- Benchmark release
- 2026-08-25
- Collected
- Not recorded
- Modality
- Not recorded
These are source-reported observations, not independent reruns by this site. Check the task version, evaluation protocol, unit and reporting date before comparing models. Results from different setups are not a universal ranking.
Original sources
Reported results 13
Shown in source observation order, not ranked. Dates retain the precision provided by the source.
| Model | Value | Unit | Protocol | Reported date and basis | Source |
|---|---|---|---|---|---|
| Apodex 1.1 | 10.3 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| Claude Opus 5 | 17.5 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| DeepSeek V4 Flash-0731 | 12.4 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| DeepSeek V4 Pro-0813 | 13.4 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| GLM-5.2 | 3.1 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| Gemini 3.7 Flash | 10.3 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| Grok 4.6 | 20.6 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| Kimi K3 | 17.5 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| Qwen 3.8 Max | 15.5 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| Qwen3.5-397B-A17B | 4.1 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Claude Code | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| GPT-5.6 Sol | 20.6 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Codex | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| GPT-5.6 Terra (max) | 15.5 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Codex | 2026-08-25 · document_publication | FrontierChallenge leaderboard |
| Apodex 1.1 | 12.4 | percent | 97 released tasks; one trajectory per system-task pair; pass = native Score >= 99.9; agent scaffold: Frontier Agent (Agent Team) | 2026-08-25 · document_publication | FrontierChallenge leaderboard |