Leaderboard wordt geladen
Anthropic · pinned anthropic/claude-opus-4.8-fast
No earlier published WFT run yet.
95% interval around 32.8 / 40, based on 197/240 answers.
Aggregated error categories only; no question text or raw model output.
| Type | Count | Rate |
|---|---|---|
| Parserfouten | 0 | 0% |
| Refusals | 0 | 0% |
| Timeouts | 0 | 0% |
| Retry releases | 0 | 0% |
| Failed jobs | 0 | 0% |
| Objective | Correct | Total | % |
|---|---|---|---|
| WFT-Basis | 197 | 240 | 82% above pass threshold |
| Advice product | Reviews | % |
|---|---|---|
| Auto | 1 | 95% above pass threshold |
| Reis | 2 | 93% above pass threshold |
| Rechtsbijstand | 1 | 90% above pass threshold |
| Woning | 3 | 90% above pass threshold |
| Zorg | 3 | 87% above pass threshold |
| AVP | 1 | 85% above pass threshold |
| Date | WFT score | Run |
|---|---|---|
| 01 Jun 2026 | 33 / 40 | combined |
These fields show which model route, parser and dataset belong to this score.
| Field | Value |
|---|---|
| Provider | Anthropic |
| Gateway | anthropic/claude-opus-4.8-fast |
| Exact API model name | anthropic/claude-opus-4.8-fast |
| Provider release date | 27 May 2026 |
| Test date | 01 Jun 2026 |
| Endpointtype | router |
| Model alias or snapshot | anthropic/claude-opus-4.8-fast |
| Provider can silently change | onbekend |
| Seed supported | No |
| Tools disabled | Yes |
| System prompt | Empty |
| Parser version |
|---|
| answer-parser-v2 |
| prompt_template_hash | 5a732017979c |
|---|
| Benchmark version | WFT-Basis v1 |
|---|
| Dataset version | InsureBench Wft-Basis v1.1 |
|---|