Agenttaken-ranglijst
Zie welke AI-modellen het best presteren op Agenttaken, welke betrouwbaar blijven en waar de grootste verschillen zitten. Sorteren op: Metriek ↑.
380/380
Modellen filteren
Geen modellen komen overeen met de huidige zoekopdracht en filters.
| Rang | Model | Bedrijf | Agenttaken-score | Score | Totale kosten | Correcte tests | Responstijd (gem.) |
|---|---|---|---|---|---|---|---|
| #277 | Ember-1 none | Fireworks | 6.1 | 5.4 | $0.786 | 0/1 | 92.9s |
| #280 | Granite 4.2 8B low | IBM Granite | 6.1 | 5.3 | $0.036 | 0/1 | 142.7s |
| #342 | gpt-oss-120b none | OpenAI | 6.1 | 4.2 | $0.016 | 0/1 | 140.9s |
| #106 | Step 3.7 Flash medium | Stepfun | 7.4 | 7.8 | $0.571 | 0/1 | 93.9s |
| #122 | GPT-5.6 Luna medium | OpenAI | 7.4 | 7.4 | $0.101 | 0/1 | 58.1s |
| #132 | GLM 5.3 Flash high | Z.ai | 7.4 | 7.3 | $0.084 | 0/1 | 111.5s |
| #147 | Step 3.7 Flash low | Stepfun | 7.4 | 7.2 | $0.449 | 0/1 | 83.6s |
| #212 | Qwen3.5-35B-A3B medium | Qwen | 7.4 | 6.3 | $1.249 | 0/1 | 126.8s |
| #222 | Trinity Large Thinking low | Arcee AI | 7.4 | 6.3 | $0.700 | 0/1 | 63.2s |
| #250 | DeepSeek V4.1 Flash none | DeepSeek | 7.4 | 5.8 | $0.123 | 0/1 | 55.8s |
| #1 | Gemini 3.6 Flash high | 10.0 | 9.9 | $1.512 | 1/1 | 65.6s | |
| #2 | Gemini 3.8 Flash high | 10.0 | 9.9 | $1.927 | 1/1 | 115.5s | |
| #3 | GPT-6 Sol high | OpenAI | 10.0 | 9.9 | $0.921 | 1/1 | 53.3s |
| #4 | GPT-6.1 Sol max | OpenAI | 10.0 | 9.9 | $1.006 | 1/1 | 80.2s |
| #5 | GPT-6 Astra max | OpenAI | 10.0 | 9.9 | $5.044 | 1/1 | 69.6s |